microbial community analysis of a uasb reactor...
TRANSCRIPT
Microbial Community Analysis of a UASB
Reactor and Application of an Evolutionary
Algorithm to Enhance Wastewater Treatment
and Biogas Production
This work is submitted in complete fulfillment for the degree of Doctor of
Philosophy (Biotechnology) in the Department of Biotechnology and Food
Technology, Faculty of Applied Sciences at the Durban University of Technology,
Durban, South Africa
Abimbola Motunrayo Enitan
(B.Sc. (Hons), M.Sc.: Microbiology)
2014
Supervisor: Dr Feroz Mahomed Swalaha
Co-supervisors: Prof. Faizal Bux, Prof. Josiah Adeyemo and Dr Sheena Kumari
ii
ABSTRACT
Anaerobic digestion, a proven and highly efficient biological process for treating industrial wastewater
and biogas generation is an underutilized technology in South Africa. Some of the industries that
have on-site anaerobic reactors tend to face problems in operating these reactors due to poor
understanding of the process and implementation of the technology. This has resulted in high
pollutant loads in their final effluents and low energy recovery. In this study, an on-site full–scale
upflow anaerobic sludge blanket (UASB) reactor treating brewery wastewater was extensively
monitored in order to evaluate the efficiency in terms of effluent quality, biogas production and
microbial structure. Furthermore, developed and adopted kinetic models were used to optimize the
performance of the full–scale UASB reactor using a combined Pareto differential evolution
(CPMDE) algorithm.
A preliminary analysis of the raw wastewater has shown that the wastewater produced from the
brewery industry was high in organic matter with a total chemical oxygen demand (COD) between
1096.41 to 8926.08 mg/L. The average removal efficiency of COD from the UASB reactor after
treatment was 79% with a methane (CH4) production of 60-69% at temperature ranges of 28-32˚C
and hydraulic retention time (HRT) of 12 h within the optimal pH range for anaerobic bacteria (6.6
and 7.3) under various organic loading rates. However, the results also showed an increase in total
suspended solids (TSS), nitrogen (N2), ammonia (NH3) and orthophosphate concentrations when
comparing the influent to the effluent, which indicated the necessity for further optimization of the
reactor condition in order to reduce these effluent parameters to acceptable standards and to increase
CH4 production.
In order to optimize the process, a thorough understanding of microbial interaction was essential. A
combination of different molecular techniques viz., fluorescence in–situ hybridization (FISH),
polymerase chain reaction (PCR) and quantitative real-time PCR (QPCR) were employed to
understand the microbial community structure of the granular sludge samples using species specific
primers and probes. The results revealed that the dominance of diverse groups of eubacteria
belonging to phyla Proteobacteria, Firmicutes and Chloroflexi and an uncultured candidate division
WS6 with four different orders of methanogenic Archaea viz., Methanomicrobiales,
iii
Methanococcales, Methanobacteriales and Methanosarcinales belonging to hydrogenotrophic and
aceticlastic methanogens were within the reactor samples. Quantification of the 16S rDNA copies of
eubacteria and methanogenic Archaea using species-specific primers further confirmed the spatial
distribution of these microorganisms within the different compartments of the reactor where, the
upper compartments were dominated by eubacteria and the lower compartments by methanogenic
Archaea. The concentration of Archaea per nanogram of DNA was much higher (96.28%) than
eubacteria (3.78%) in lower compartments, while, the eubacteria concentration increased to 98.34%
in upper compartments with a decrease in Archaea quantity (1.66%).
A modified kinetic methane generation model (MMGM) was developed on the basis of mass balance
principles with respect to substrate (COD) degradation and the endogenous decay rate to predict CH4
production efficiency of the reactor. Furthermore, a Stover–Kincannon kinetic model was adopted
with the aim of predicting the final effluent quality in terms of COD concentration and model
coefficients were determined using the data collected from the full–scale reactor. Thereafter, a
model-based multi-objective optimization was carried out using the CPMDE algorithm with three–
objective functions namely; maximization of volumetric CH4 production rate; minimization of
effluent substrate concentration and minimization of biomass washout, in order to increase the
overall efficiency of the UASB reactor. Important decision variables and constraints related to the
process were set for the optimization. A set of non-dominated solutions with high CH4 production
rates of between 2.78 and 5.06 L CH4/g COD/day at low biomass washout concentrations were
obtained at almost constant solution for the effluent COD concentration. A high COD removal
efficiency (85-87%) at ~30-31˚C and 8-9 h HRT was obtained for the multi-objective optimization
problem formulated.
This study could significantly contribute towards optimization of a full–scale UASB reactor treating
brewery wastewater for better effluent quality and biogas production. Knowledge on the activity and
performance of microbial community present in the granular sludge taken from the full–scale UASB
reactor would contribute significantly to future optimization strategies of the reactor. In addition,
optimization using an evolutionary algorithm under different operational conditions would help to
save both time and resources wasted in operating anaerobic bioreactors.
iv
DECLARATION
―I declare that the thesis herewith submitted for the degree Doctor of Philosophy: Biotechnology at
the Durban University of Technology is my original work and has not been previously submitted for
a degree at any other institution of higher education, and that its only prior publication was in the
form of conference papers, book chapter and/or journal articles. I further declare that all the sources
cited or quoted are acknowledged and indicated by means of a comprehensive list of references‖.
A.M. Enitan
I hereby approve the final submission of the following thesis.
Dr F. M. Swalaha Prof. F. Bux Prof. J. Adeyemo Dr Sheena Kumari
D. Tech. (DUT) D. Tech. (DUT) D. Tech. (TUT) PhD (Mangalore University)
v
DEDICATION
I am dedicating this project to Jehovah, the father of the whole Universe who made this project a
reality. To the memories of my brother Enitan, Ibukunoluwa Olabisi—my friend and brother–an
encourager and motivator, who passed on without witnessing the results of his pieces of advice
and motivation. We shall meet again in Paradise.
vi
ACKNOWLEDGEMENTS
I would like to express my sincere appreciation to Jehovah, Sovereign Lord of the Universe, the
provider of knowledge, wisdom and understanding for making this dream come true.
Dr Feroz Mahomed Swalaha for his supervision, advice, patience and open door policy
throughout the course of this study. I am very grateful for his genuine disposition at various
stages of this study.
I am grateful to my co-supervisor, Professor Faizal Bux for his supervision, time and for been a
good host in providing financial support as well as equipment needed for this project.
I am very grateful to my co-supervisor, Professor Josiah Adeyemo for his unwavering support,
encouragement, guidance, assistance and being inspirational. I am very grateful for his critical
suggestions and constructive criticism.
I enjoyed the good working relationship with Dr Sheena Kumari. I am very grateful for her time,
believing in me, dedication and supervision throughout the course of this study.
Mr. Oluwatosin Olofintoye, Achisa Cleophas and Jaafar Bux for their constant help and for
teaching me how to use the algorithm, as well as the engineering aspect of this work.
I am deeply grateful to my guidance, Mr. and Mrs. Ipadeola for their Godly training, fatherly and
motherly love, encouragement and dedication to make this degree a success. I would like to
thank them for their prayers and supports, I would forever be grateful for given me life and
showing interest in me.
I am saying big thank you to my siblings; Engr. Temitayo and Ronke Enitan, Lillian Enitan and
my cousins. Thank you all for your love, encouragements and for standing by me through the
hard and good times of my life and throughout the course of this study. Thank you to Dr and
Mrs. Olusola Olubiyi for their constant supports and encouragements.
I would like to acknowledge Mr. Luis Lucamba, Mr. Agbejoye Durotimi and Saheed Oladiti for
their love and support for me during the course of this study and my friends who were constant
source of inspiration to me. Thank you all.
vii
I would like to acknowledge my special friends; Dr Durotolu Amosun, Dr Benjamin Okeleye, Dr
and Mrs. Ojodu, Dr Shade Adeyinka and Mr. Oyewole Stanley for their encouragements and
supports towards the success of this degree.
Special thanks to my colleagues and staff of Institute for Water and Wastewater Technology, for
their supports and friendship, Mr. Oluyemi Awolusi, Dr Nishani, Thobela Conco and Kriveshin
Pillay to mention few, thank you all.
My appreciation goes to Durban University of Technology for the financial support and
scholarship to pursue this degree, thereby making this dream possible. I would always be grateful
to this institution.
Special thanks to the Post Graduate Development Support office at the Durban University of
Technology for funding conference attendances.
viii
TABLE OF CONTENTS
ABSTRACT .................................................................................................................................. II
DECLARATION......................................................................................................................... IV
DEDICATION.............................................................................................................................. V
ACKNOWLEDGEMENTS ....................................................................................................... VI
TABLE OF CONTENTS ........................................................................................................ VIII
LIST OF FIGURES ................................................................................................................. XIII
LIST OF TABLES ................................................................................................................... XVI
ABBREVIATIONS ................................................................................................................ XVII
PREFACE ................................................................................................................................. XIX
CHAPTER ONE: INTRODUCTION ......................................................................................... 1
1.1 STUDY OBJECTIVES ..................................................................................................... 5
1.1.1 Aim ............................................................................................................................... 5
1.1.2 Objectives ..................................................................................................................... 5
1.2 THESIS OUTLINE ........................................................................................................... 6
CHAPTER TWO: LITERATURE REVIEW ............................................................................ 7
2.1 INTRODUCTION............................................................................................................. 7
2.2 ANAEROBIC TREATMENT OF WASTEWATER ..................................................... 8
2.2.1 Upflow Anaerobic Sludge Blanket Reactors ................................................................ 9
2.3 BIOGAS RECOVERY FROM ANAEROBIC DIGESTERS ..................................... 11
2.4 BIOCHEMISTRY AND MICROBIOLOGY OF THE ANAEROBIC DIGESTION
PROCESS ........................................................................................................................ 13
2.4.1 Hydrolytic Bacteria ..................................................................................................... 14
2.4.2 Fermentative Acidogenic Bacteria ............................................................................. 15
2.4.3 Acetogenic Bacteria .................................................................................................... 15
2.4.4 Methanogenic Archaea and their Taxonomy.............................................................. 16
ix
2.4.5 Techniques Used To Detect Microorganisms from Anaerobic Reactor Samples ...... 20
2.5 FACTORS AFFECTING PERFORMANCE OF UASB REACTORS AND BIOGAS
PRODUCTION ............................................................................................................... 24
2.5.1 Organic Loading Rate ................................................................................................. 24
2.5.2 Nutrients ..................................................................................................................... 24
2.5.3 Hydraulic Retention Time .......................................................................................... 25
2.5.4 Volatile Fatty Acids .................................................................................................... 26
2.5.5 Operational Temperature ............................................................................................ 26
2.5.6 Operational pH ........................................................................................................... 27
2.6 MODELLING OF ANAEROBIC DIGESTION SYSTEMS ....................................... 28
2.7 OPTIMIZATION TECHNIQUES USING EVOLUTIONARY ALGORITHMS .... 31
2.8 RESEARCH OUTPUT ................................................................................................... 38
CHAPTER THREE: PERFORMANCE EVALUATION OF AN UPFLOW ANAEROBIC
SLUDGE BLANKET REACTOR TREATING BREWERY WASTEWATER .................. 39
3.1 INTRODUCTION........................................................................................................... 39
3.2 MATERIALS AND METHODS .................................................................................... 41
3.2.1 Description of Full-Scale UASB Reactor ................................................................... 41
3.2.2 Wastewater and Biogas Sampling Procedure ............................................................. 42
3.2.3 Wastewater Characterization ..................................................................................... 42
3.2.3.1 Conventional and instrumental methods used for analysis ............................... 43
3.2.4 Analytical Quality Assurance and Statistical Analysis .............................................. 45
3.2.5 Estimation of Pollutant Removal Efficiency .............................................................. 46
3.3 RESULTS AND DISCUSSION ...................................................................................... 46
3.3.1 Brewery Wastewater Composition ............................................................................. 46
3.3.2 Efficiency of UASB Reactor Treating Brewery Wastewater ..................................... 48
3.3.2.1 Effect of pH and temperature on UASB reactor performance .......................... 48
3.3.2.2 COD removal efficiency and solids concentration ............................................ 51
3.3.2.3 Nitrogen and phosphate concentrations in the wastewater ............................... 54
3.3.2.4 Correlation between methane production and operational variables ................ 56
3.4 CONCLUSIONS ............................................................................................................. 60
3.5 RESEARCH OUTPUTS ................................................................................................. 61
x
CHAPTER FOUR: KINETIC MODELLING AND CHARACTERIZATION OF THE
MICROBIAL COMMUNITY PRESENT IN AN UASB REACTOR TREATING
BREWERY EFFLUENT ........................................................................................................... 62
4.1 INTRODUCTION........................................................................................................... 62
4.2 MATERIALS AND METHODS .................................................................................... 64
4.2.1 Sample Collection from the Full-Scale UASB Reactor ............................................. 64
4.2.2 Fluorescence In-Situ Hybridization (FISH) ................................................................ 65
4.2.2.1 Microscopy and image analysis ......................................................................... 65
4.2.3 Total Genomic DNA Extraction from Granular Sludge Samples .............................. 66
4.2.4 Amplifications using Polymerase Chain Reaction (PCR) .......................................... 67
4.2.4.1 Agarose gel electrophoretic detection of PCR products ..................................67
4.2.4.2 Cloning .............................................................................................................. 68
4.2.4.2.1 Preparation of competent cells, ligation, transformation and clone
analysis using colony PCR ..................................................................................... 68
4.2.4.3 Sequencing and phylogenetic analysis............................................................. 69
4.2.4.3.1 Nucleotide sequence accession number for samples obtained from the
full-scale UASB reactor .......................................................................................... 69
4.2.5 Quantitative Real-time PCR ....................................................................................... 70
4.2.6 Kinetic Analysis Using Stover–Kincannon Model .................................................... 71
4.2.7 Statistical Analysis ..................................................................................................... 73
4.3 RESULTS AND DISCUSSION ...................................................................................... 73
4.3.1 Profiling of Microbial Community Structure of a Full-Scale UASB Reactor Granules
Based on 16S rDNA Analysis................................................................................................. 73
4.3.1.1 Characteristics of granular sludge used for the molecular analysis .................. 73
4.3.1.2 Methanogenic Archaea and bacteria detected from the granular sludge using
FISH technique ................................................................................................................ 74
4.3.2 Community of the Granular Sludge Using PCR ........................................................ 76
4.3.2.1 Bacterial diversity within the reactor compartments ....................................... 76
xi
4.3.2.2 Archaea composition in the granular sludge ..................................................... 82
4.3.2.3 Detection of methyl coenzyme-M reductase gene A (mcrA) in the granular
sludge .............................................................................................................................. 86
4.3.3 Optimization of QPCR for Quantification of Microbial Communities Present in the
Granular Sludge Samples ........................................................................................................ 89
4.3.3.1 Comparison of concentration of Archaea and bacterial communities in the
different reactor compartments ....................................................................................... 90
4.3.4 Performance of UASB Reactor and Biogas Production ............................................. 94
4.3.5 Kinetic Modelling and Model Validation ................................................................... 97
4.4 CONCLUSIONS ........................................................................................................... 100
4.5 RESEARCH OUTPUTS ............................................................................................... 101
CHAPTER FIVE: DEVELOPMENT OF A MATHEMATICAL MODEL TO DESCRIBE
THE BEHAVIOUR AND PERFORMANCE OF A UASB REACTOR TREATING
BREWERY WASTEWATER FOR BIOGAS PRODUCTION ........................................... 102
5.1 INTRODUCTION ........................................................................................................ 102
5.2 MATERIALS AND METHODS .................................................................................. 105
5.2.1 Ghaly et al. (2000) Model ........................................................................................ 105
5.2.1.1 The microbial mass balance ............................................................................. 105
5.2.1.2 Substrate mass balance and effluent substrate concentration ............................ 107
5.2.1.3 Biogas production .............................................................................................. 108
5.2.2 Modified Methane Generation Model (MMGM) ..................................................... 109
5.2.3 Determination of MMGM Parameters (K, µmax, Kd , Y and Bo) ............................... 112
5.2.4 Software Used and Statistical Analysis .................................................................... 113
5.2.5 Description of the UASB Reactor System Used and Wastewater Sampling ........... 113
5.2.6 Calculation of Methane Potential and Yield (United Nations Economic Commission
for Europe, 2004) .................................................................................................................. 114
5.3 RESULTS AND DISCUSSION .................................................................................... 114
5.3.1 Estimated MMGM Parameters ................................................................................. 114
5.3.2 Validation of the Modified Methane Generation Model .......................................... 119
5.4 CONCLUSIONS ........................................................................................................... 124
5.5 RESEARCH OUTPUT ................................................................................................. 125
xii
CHAPTER SIX: MULTI-OBJECTIVE OPTIMIZATION OF A METHANE–
PRODUCING UASB REACTOR USING A COMBINED PARETO MULTI-OBJECTIVE
DIFFERENTIAL EVOLUTION ALGORITHM .................................................................. 126
6.1 INTRODUCTION......................................................................................................... 126
6.2 METHODS .................................................................................................................... 129
6.2.1 Optimization of UASB Reactor ............................................................................... 129
6.2.2 Combined Pareto Multi-Objective Differential Evolution (CPMDE) Algorithm .... 132
6.2.2.1 The CPMDE algorithm .................................................................................... 132
6.2.2.2 Implementation of CPMDE algorithm for optimization of an UASB reactor . 134
6.3 RESULTS AND DISCUSSION .................................................................................... 134
6.4 CONCLUSIONS ........................................................................................................... 139
6.5 RESEARCH OUTPUT ................................................................................................. 140
CHAPTER SEVEN: GENERAL CONCLUSIONS AND RECOMMENDATIONS ......... 141
7.1 SIGNIFICANCE AND NOVELTY OF THE RESEARCH FINDINGS ................. 144
7.2 RECOMMENDATIONS .............................................................................................. 145
REFERENCES .......................................................................................................................... 147
APPENDICES ........................................................................................................................... 184
This thesis is a compilation of different manuscripts, where each chapter is an individual entity and
some repetition is unavoidable between chapters.
xiii
LIST OF FIGURES
Figure 2.1: (A) Proportions of the types of developed anaerobic digestion systems that have
been installed and commercialized for the treatment of industrial wastewater (International
Energy Agency, 2001); (b) percentage of industries using anaerobic treatment technologies for
industrial wastewater. ......................................................................................................................9
Figure 2.2: Schematic diagram of an upflow anaerobic sludge bed (UASB) reactor with red balls
indicating granules and yellow balls indicating evolved biogas. ...................................................10
Figure 2.3: The key stages of anaerobic digestion of organic matter in the wastewater (Li et al.,
2011). .............................................................................................................................................14
Figure 2.4: Classification of methanogens based on 18S and 16S rRNA analysis and comparison
of conservative phylogenetic features (Demirel and Scherer, 2008b; Ziemiński and Frąc, 2012).17
Figure 2.5: Pathways of methanogenesis: hydrogenotrophic (double-lined arrows), aceticlastic
(solid arrows) and methylotrophic (broken gray arrows) (Bapteste et al., 2005). .........................19
Figure 2.6: Flow diagram of different steps used in studying the structure and functions of
microbial communities in an environmental samples. ...................................................................23
Figure 2.7: Flowchart for evolutionary algorithm development. ...................................................32
Figure 2.8: Flowchart for the main steps in DE algorithm development. ......................................34
Figure 3.1: Layout of full-scale UASB reactor treating brewery wastewater (Hoffmann, 1985;
Ross, 1989). ...................................................................................................................................44
Figure 3.2: Schematic diagram of the sampling points from which samples were collected for
this study to monitor the full-scale UASB reactor treating brewery wastewater. .........................45
Figure 3.3: The effect of inlet COD variations on the pH of the full-scale UASB reactor treating
brewery wastewater. ......................................................................................................................49
Figure 3.4: (a) Change and (b) the relationship between reactor temperature and final pH value
of UASB reactor treating brewery wastewater. .............................................................................50
Figure 3.5: Performance of the full-scale UASB reactor treating brewery wastewater in terms of
COD removal efficiency. ...............................................................................................................51
Figure 3.6: (a) Performance of the UASB reactor treating brewery wastewater in terms of total
suspended solids removal and (b) the second order quadratic polynomial regression between
%TSS and %COD removal efficiency of the UASB reactor… .....................................................53
Figure 3.7: Variation in average inlet and outlet concentrations of ammonia nitrogen during
anaerobic treatment of brewery wastewater using UASB reactor. ................................................55
Figure 3.8: Average orthophosphate concentration in the reactor during treatment of brewery
wastewater......................................................................................................................................56
xiv
Figure 3.9: Efficiency of organic matter removal (COD quantity) as function of reactor volume
to produce biogas during anaerobic treatment of brewery wastewater ..........................................57
Figure 3.10: Effect of organic loading rate on methane production rate in a UASB reactor
treating brewery wastewater. .........................................................................................................58
Figure 3.11: Graph showing (a) the effect of reactor‘s pH on the methane content and, (b) the
relationship and linear regression analysis showing a significant negative correlation between
these two parameters during the treatment of brewery wastewater using UASB
reactor…………………………………………………………………….……………………...59
Figure 4.1: Flow diagram showing the six sampling points from the UASB reactor compartments
where granular samples were obtained for microbial
analysis..………………………………………………………………………………………… 64
Figure 4.2: (a) Images of granules hybridized by highly rhodomine labeled archaeal-domain
oligonucleotide probes (ARC915) showing diverse species of methanogens (green); (b)
corresponding image of ARC915 granules showing diverse species of methanogens stained with
DAPI (blue), (c) granular sludge of FISH labeled with tetramethylrhodomine-5-isothiocyanate
using the universal probes for eubacteria (EUB338), (d) the MX825 probe labeled sample to
confirmed the acetoclastic Methanosaeta group and (e) the corresponding DAPI stained cells for
EUB mix……………………………………………………………………………………. .......75
Figure 4.3: Agarose gel depicting PCR products for the bacterial fragments (1500 bp). The bands
corresponding to lanes C1–C6 represent the bacterial fragments from the six compartments of
the UASB reactor when PCR amplification was performed using 27f/1492r specific primer set.
Lane L corresponds to the 1 kb DNA marker used in this study…………………………… .......77
Figure 4.4: Phylogenetic tree of bacterial clones obtained from granular sludge of UASB reactor
treating brewery wastewater using universal 27f/1492r bacterial primer set. The evolutionary
history was inferred using the neighbor-joining method (Saitou and Nei, 1987). The nucleotide
sequences were submitted to the National Centre for Biotechnology Information website under
the accession numbers KM065733 – KM065740 corresponding to the selected clones (1B-10B)
from compartments C1, C3 and C6………………..……………………………………….. .......79
Figure 4.5: Agarose gel showing 16S rDNA gene PCR fragments obtained from the
amplification of genomic DNA extracted from the granular sludge samples using ARC primer
set. Bands corresponding to lanes C1–C6 represent the Archaea fragments from the six
compartments of the UASB reactor between 243–250 bp using 1 kb DNA marker (Lane L) in the
analysis……………………………………………………………………………………….. .....83
Figure 4.6: Phylogenetic tree for methanogenic Archaea obtained from granular sludge of UASB
reactor treating brewery wastewater using methyl coenzyme-M reductase (mcrA) gene primer
set. The evolutionary history was inferred using the neighbor-joining method (Saitou and Nei,
1987). The GenBank accession numbers are KF715644–KF715648 corresponding to the selected
clones…………………………………………………………………………………….…. .......88
xv
Figure 4.7: Variation in the percentage of bacteria and Archaea communities in the granules
collected at the different reactor compartments (C1–C6) using universal primer sets for the
quantitative real-time PCR assay, in this study……………………………...……………….. ....91
Figure 4.8: Abundance of Archaea and bacterial DNA copy numbers of 16S rDNA genes per
nanogram of genomic DNA extracted from the granular samples obtained from each
compartments of the full-scale UASB reactor using QPCR assays for the primer sets used in this
study.……………………………………………………………………………………...… .......94
Figure 4.9: Effect of organic loading rate on COD removal rate using the modified Stover-
Kincannon model to determine the kinetic constants……………………………………….. ......99
Figure 4.10: Relationship between the observed and predicted effluent COD concentrations by
modified Stover-Kincannon model……………………………………………….…………. ....100
Figure 5.1: Schematic diagram of a single compartment of an upflow anaerobic sludge blanket
reactor (see abbreviations for definition of symbols)………………………………………. .....106
Figure 5.2: The time–course of COD and BOD5 removal efficiencies for the full–scale UASB
reactor treating brewery wastewater over the period of time, in this study.………………... .....115
Figure 5.3: Estimation of the kinetic parameter K and the maximum growth rate of
microorganism‘s µmax, from data collected from the full–scale UASB reactor treating brewery
wastewater. The plot of θh against S [where, S = (Si–Se/Se)] gives a straight line with 1/intercept
as µmax and slope/intercept as K………………………………………………………….…. .....117
Figure 5.4: Ultimate methane yield (Bo) obtained from data collected from the full–scale UASB
reactor treating brewery wastewater by plotting methane yield against the reciprocal of hydraulic
retention time…………………………………………………………….………………….. ....118
Figure 5.5: The endogenous decay coefficient, Kd and the growth yield coefficient, Y were
calculated from the intercept and slope of the straight line of the plotted graph using the data
obtained from the full–scale UASB reactor treating brewery wastewater………………..... .....118
Figure 5.6: Observed and predicted methane yields at different hydraulic retention times… ....120
Figure 5.7: (a) The trend between observed and predicted volumetric methane production rates at
different organic loading rates using the newly developed model and (b) the scatter plot of
predicted vs observed volumetric methane production rates relationship between them at lower
organic loading rates.……………….…………………..……………………………………. ...122
Figure 5.8: The predicted and observed volumetric methane production rates at different
temperatures using the developed model (MMGM) ……….………………………………. .....124
Figure 6.1: Pareto optimal set of solutions obtained for the simultaneous optimization of
volumetric methane production rate (Yv), effluent biomass concentration (Xe) and effluent
substrate concentration (Se) as a multi–objective optimization problem..……………………. ..136
Figure 6.2: The Optimal decision variables (a) θh and (b) P plotted against volumetric methane
production rate (Yv), as well as (c) θh and (d) P plotted against effluent biomass concentration
(Xe) for the optimized problem……….…………………………………………………..….. ...138
xvi
LIST OF TABLES
Table 2.1: Optimum pH ranges for selected methanogens (Gerardi, 2003; Steinhaus et al., 2007).27
Table 2.2: Anaerobic model and optimization tools for different types of wastewater ............... 37
Table 3.1: Summary of raw brewery wastewater composition from the industry prior to
anaerobic treatment and indicative discharge limits in South Africa (SA) and the EU ...............47
Table 3.2: Brewery wastewater characterization and the efficiency of the UASB reactor as
compared to the literature ..............................................................................................................48
Table 3.3: Composition of influent (brewery wastewater after pre-conditioning) and UASB
effluent ......................................................................................................................................... 52
Table 4.1: 16S rRNA oligonucleotide probes with the corresponding formamide stringency and
NaCl concentrations used in this study ......................................................................................... 66
Table 4.2: Primer sets used in this study for both conventional and quantitative real-time PCR 69
Table 4.3: Characterization of granular sludge used for molecular analysis ................................ 72
Table 4.4: Bacterial community profiles of the clones retrieved from granular sludge samples
taken from the UASB reactor, as compared to the known sequences in the GenBank database 78
Table 4.5: Sequence similarity of Archaea from the full-scale UASB reactor with the GenBank
database sequences....................................................................................................................... 84
Table 4.6: Description of QPCR standard curves parameters for 16S rDNA copy number for
ARC as the universal Archaea and BAC as the universal bacterial primer sets that are responsible
for biological conversion of complex organic matter in the brewery wastewater into simple
monomer and CH4 production ...................................................................................................... 90
Table 4.7: Biochemical properties of pre-conditioned brewery wastewater entering the UASB
reactor before treatment ................................................................................................................ 95
Table 4.8: Average compositions of biogas produced in this study ............................................. 95
Table 4.9: Comparison of different types of anaerobic wastewater treatment processes using the
modified Stover–Kincannon model .............................................................................................. 98
Table 5.1: Average data obtained from the full-scale UASB reactor treating brewery wastewater
..................................................................................................................................................... 116
Table 5.2: Data used for the determination of MMGM parameters ........................................... 116
Table 5.3: Estimated MMGM parameters as obtained using the data collected from the full–scale
UASB reactor treating brewery wastewater ............................................................................... 116
Table 5.4: Kinetic parameters obtained in this study compared to other studies ....................... 119
Table 6.1: Details of model-based multi-objective optimization problem studied using CPMDE
algorithm ......................................................................................................................................131
Table 6.2: The CPMDE parameters used for multi-objective optimization problem .................135
xvii
ABBREVIATIONS
AD : anaerobic digestion
ANN : artificial neural network
ANN-GA: artificial neural network coupled with genetic algorithm
b : dimensionless kinetic parameter
B : Actual volume of methane produced (in litres) per gram of COD
(substrate) added to the reactor at S.T.P.
Bo : ultimate methane yield coefficient under normal conditions of temperature and
pressure per gram of substrate (COD) added for complete utilization of substrate or
at an infinite hydraulic retention time
BOD : biological oxygen demand
CH4 : methane
CO2 : carbon dioxide
COD : chemical oxygen demand
CPMDE : Combined Pareto Multi-Objective Differential Evolution
Cq : quantification cycle
: rate of substrate removal, (g/ L/ d)
: rate of change in microbial mass, (g/ L/ d)
DE : differential evolution
EA : evolutionary algorithm
GA : genetic algorithm
HRT : Hydraulic retention time
gDNA : genomic deoxyribonucleic acid
K : biokinetic constant
Kd : endogenous decay coefficient, (/d)
MMGM : modified methane generation model
ng : nanogram
NH3 : ammonia
OLR : organic loading rate
P : fraction of biodegradable COD,
Q : flow rate, (L /d)
S : concentration of substrate, (g COD/L)
Se : effluent substrate concentration, (g/ L)
Si : influent substrate concentration, (g/ L)
Sr : concentration of substrate in the reactor, (g/ L)
T : operational temperature, (◦C)
TS : total solid
TSS : total suspended solid
xviii
TVS : total volatile solid
UASB : upflow anaerobic sludge bed
VFA : volatile fatty acid
VS : volatile solid
VSS : volatile suspended solid
Vr : reactor volume, (L)
X : microbial cell concentration, (g / L)
Xe : concentration of biomass in the effluent, (g/L)
Xi : concentration of biomass in the influent, (g/L)
Xr : concentration of biomass in the reactor, (g/L)
Y : growth yield coefficient, (g/g)
Yv : volumetric methane production rate, (L methane/g COD added/d)
θh : hydraulic retention time, (/Time),
µmax : maximum growth rate of microorganisms when the substrate is being
used at its maximum rate
µ : specific growth rate of microorganisms, (/d1)
xix
PREFACE
PUBLICATIONS
This work has resulted in the following publications.
(a) Book Chapter
1) Enitan, A. M., Adeyemo, J., Olofintoye, O. O., Bux, F. and Swalaha, F. M. (2014). Multi-
objective optimization of a methane producing UASB reactor using a combined Pareto multi-
objective differential evolution algorithm. EVOLVE - A Bridge between Probability, Set
Oriented Numerics, and Evolutionary Computation V. Advances in Intelligent Systems and
Computing, Springer, 288: 321-334.
(b) Journal Articles
1) Enitan, A. M., Kumari, S., Swalaha, F. M., Adeyemo, J., Ramdhani, N. and Bux, F. (2014).
Kinetic modelling and characterization of microbial community present in a full-scale UASB
reactor treating brewery effluent. Microbial Ecology, 67: 358–368.
2) Enitan, A. M., Swalaha, F. M., Adeyemo, J. and Bux, F. (2014). Assessment of brewery
effluent composition from a beer producing industry in KwaZulu-Natal, South Africa.
Fresenius Environmental Bulletin, 23 (3): 693-701.
3) Adeyemo, J. and Enitan, A. (2011). Optimization of fermentation processes using
evolutionary algorithms. Scientific Research and Essays, 6 (7): 1464-1472.
4) Enitan, A. M. and Adeyemo, J. (2011). Food processing optimization using evolutionary
algorithms. African Journal of Biotechnology, 10 (72): 16120-16127.
xx
(c) Conference Papers
1) Enitan, A. M., Swalaha, F. M., Adeyemo, J. and Bux, F. (2014). Evaluation of effluent
composition from a beer producing industry in South Africa. Presented at the International
Journal of Arts & Sciences’ (IJAS) American Canadian Conference at Ryerson University’s
International Learning Center, Toronto, Canada, 19-22 May, 2014 (Oral presentation).
2) Enitan, A. M., Kumari, S., Swalaha, F. M. and Bux, F. (2014). Real-time PCR for
quantification of methanogenic Archaea in a UASB reactor treating brewery wastewater.
Presented at International Journal of Arts & Sciences’ (IJAS) American Canadian
Conference at Ryerson University’s International Learning Center, Toronto, Canada, 19-22
May, 2014. Conference of the International Journal of Arts & Sciences, CD-ROM. ISSN:
1943-6114 : 07(03):103–106.
3) Adeyemo, J. and Enitan, A. M. (2014). Multi-objective optimization of anaerobic digestion
models for biogas production. Presented at International Journal of Arts & Sciences’ (IJAS)
for academic disciplines Conference at Harvard Medical School, 77 Louis Pasteur, Boston,
Massachusetts, 26-30 May, 2014 (Oral presentation).
4) Enitan, A. M., Kumari, S., Swalaha, F. M., and Bux F. (2014). Use of mcrA-targeted real-
time quantitative PCR for quantification of methanogenic communities in reactor treating
brewery wastewater. Presented at Water Institute of Southern Africa (WISA) Conference,
Mbombela, Mpumalanga, South Africa, May 25-29, 2014 (Oral presentation).
5) Swalaha F. M., Enitan A. M. and Bux, F. (2014). Efficiency of industrial scale anaerobic
reactor treating brewery wastewater. Presented at Water Institute of Southern Africa (WISA)
Conference, Mbombela, Mpumalanga, South Africa, May 25-29, 2014 (Oral presentation).
6) Enitan, A. M. and Adeyemo, J. (2014). Estimation of bio-kinetic coefficients for treatment
of brewery wastewater. Presented at the World Academy of Science, Engineering and
Technology Conference, New York, USA, June 5-6 . International Science Index, 8(6): 365-
369.
1
CHAPTER ONE: INTRODUCTION
Industries produce millions of cubic meters of wastewater every year. The wastewater
produced may be released into the surrounding rivers, or treated on site or at municipal
treatment plants. With competing demand for water resources and water reuse, appropriate
discharge of industrial effluents into the aquatic environment has become an important issue,
which has led to considerable public debate (Phiri et al., 2005; Baig et al., 2010; Danazumi
and Hassan, 2010; Bello-Osagie and Omoruyi, 2012). Some industries have been fined by
their national water and municipal authorities for discharging poor quality effluents that do
not meet the discharge standards into natural water bodies as well as municipal wastewater
treatment plants (Ikhu-Omoregbe et al., 2005; Phiri et al., 2005). Also, much attention has
been placed on the impact of industrial wastewater on domestic wastewater treatment plants
and water bodies worldwide due to accumulation of organic and inorganic compounds in the
water bodies (Islam et al., 2006; Kanu and Achi, 2011; Kovoor et al., 2012).
Brewery industries, among others, produce millions of litres of various types of beers each
year with global beer production in 2011 estimated to be about 192.71 million kilolitres of
beer (Kirin Holdings, 2012) with an average consumption of 23 litres per person per year
(Fillaudeau et al., 2006). As large volumes of water are being used by the industries in the
production of beer, the amount of wastewater that is being discharged from the industries
after production is very high in organic content and thus highly polluting to the environment
(Jones et al., 2011).
Anaerobic digestion (AD) technology has long been used for the treatment of industrial
wastewater. It is a complex biological process that has been adopted for effective treatment of
organic wastewater in the absence of oxygen by microorganisms. It is used not only as a
pollution control tool, but also for energy recovery (Tiwari et al., 2006; Stafford et al., 2013).
The success of AD technology in the removal of high chemical oxygen demand (COD) has
led to its increasing application in treating many types of wastewater for bioconversion of
organic matter to biogas and better effluent quality (Castillo et al.,1999; Liu et al., 2003;
Bhunia and Ghangrekar, 2008). The development of the upflow anaerobic sludge blanket
(UASB) reactor, initially by Lettinga et al. (1980), has made AD the most competitive and
2
favoured treatment technology to process industrial wastewater in some parts of the world (Li
et al., 2014).
An UASB reactor is a biogas-producing digester that uses complex and sequential
biochemical processes through the association of anaerobic microorganism (Lettinga and
Hulshoff-Pol, 1991; Tiwari et al., 2006). The methane content of biogas produced is known
as an environmentally friendly, clean fuel which is part of nature‘s own cycle that can be
used for lighting, cooking, and running internal combustion engines. Thus, the use of
anaerobic waste fermentation to produce biogas is a promising and economical way of
generating renewable energy at the industrial scale. In recent years, UASB reactors have been
successfully applied to the treatment of different types of wastewater for better effluent
quality and in turn to produce biogas as a source of renewable energy (Cronin and Lo, 1998;
Parawira et al., 2005; Manhokwe et al., 2009; Madukasi and Zhang, 2010; Muda et al., 2011;
Nacheva et al., 2011). Anaerobic breakdown of organic compounds to biogas involves the
action of several groups of microorganisms (hydrolytic, acidogenic, acetogenic and
methanogenic bacteria) that grow in a syntrophic manner when the reactor is operated under
optimum reaction conditions (Hulshoff-Pol et al., 2004; Crocetti et al., 2006; Mumme et al.,
2010; Amani et al., 2011). Studies have shown that the microbial community in the UASB
reactor responds to the changes in environmental and operational conditions (Klocke et al.,
2007; Khalid et al., 2011; Ziemiński and Frąc, 2012; Enitan et al,. 2014b; Jang et al., 2014).
Thus, in order to optimize the performance of a UASB reactor, it is necessary to identify and
quantify the microbial communities for better treatment efficiency and biogas production
(Chen et al., 2008). Effluent quality and the amount of biogas produced depend on the type of
substrate, digester configuration and the environmental conditions (Traversi et al., 2014).
Therefore, to understand and predict the phenomena occurring in AD processes, increase the
plant performance and methane (CH4) production, holistic mathematical models are required.
To this effect, different AD models have been developed to describe and predict increased
treatment efficiency and optimize the operating conditions of the digestion system (Batstone
et al., 2002; Parawira et al., 2005; Parsamehr, 2012). Mathematical modelling is more
effective in providing information on the interactive behaviour of various factors in
fermentation processes compared to conventional one-at-a-time-optimization methods, which
3
are less effective at representing the interactive effects of all the factors involved in a
complex bioprocess (Lakshmi et al., 2009).
Mathematical modelling is useful for designing, predicting and controlling anaerobic
processes. It can assist process engineers to design new configuration of reactors for higher
efficiency and to improve the efficiency of an existing system. A process model can be either
mechanistic or empirical (Thorin et al., 2012; Estes et al., 2013). Simple and sophisticated
models for several systems of AD processes have been developed to fulfill the increasing
need of understanding the parameters required to improve the efficiency of bioreactors
(Batstone et al., 2002; Parsamehr, 2012). Technical approaches for the development of these
dynamic models to adequately describe the treatment processes and biogas production vary
from one method to another. However, the integration of different parameters, linear and
nonlinear equations with single and multi-objective functions under different constraints in
large-scale engineering problems have contributed to the development of alternative solutions
(Babu et al., 2005; Iqbal and Guria, 2009, Abu Qdais et al., 2010). A new optimization and
control strategy with immense benefits to manage AD of wastewater for energy generation
and production of better effluent that may reduce pollution to the environment is essential for
effective reactor operation.
Optimization can be defined as the art of finding one or more feasible solutions
corresponding to extreme values of one or more objectives problems, while satisfying
specific constraints (Babu et al., 2005). Optimization problems are divided into two, namely
single- and multi-objective optimization (Fister et al., 2013). The single–objective
optimization problem involves one objective function to which heuristic-based and gradient-
based search techniques are applied in order to solve the single–objective optimization
problem. It involves finding the minimum and maximum of a single–variable function. The
single–objective optimization method is employed for finding an optimum of a first– and
second–order derivative of a function. It may also involve finding the true optimum in the
presence of constraints to get solutions to real world problems (Adeyemo and Otieno, 2009a).
4
On the other hand, multi–objective optimization problem (MOOP) is an optimization
problem solving method that has more than one objective functions. It involves finding one
or more optimum solutions to more than one objective optimization problems that are
conflicting in nature (Deb, 2011). The aim of MOOP is to simultaneously optimize a set of
conflicting objectives to obtain a group of alternative trade-off solutions called Pareto-
optimal or non-inferior solutions which must be considered equivalent in the absence of
specialized information concerning the relative importance of the objectives (Deb, 2011).
With regards to all objectives, there is no best solution rather; the solutions are equally good
solutions. Meanwhile, most real-world search and optimization problems are multi-objectives
in nature with all the objective functions being very important (Fister et al., 2013).
However, limited knowledge with highly complex and non-linear digestion processes was
one of the underlying problems in AD due to lack of an online-measurements for most of the
industrial biogas–producing plants. This problem has led to the development of new
optimization and control strategies with respect to external influences and different process
disturbances that are vital for efficient operation of AD treatment plants (Sendrescu, 2013).
One approach to address this problem is to exploit the flexibility and power of computational
intelligence of evolutionary algorithms (EAs).
Evolutionary algorithms as a class of direct search algorithms have proven to be an important
tool to solve optimization problems and thus, have been employed more often during the last
decade due to their ease way of handling multiple-objective problems (Woldesenbet et al.,
2009). Constrained or unconstrained multi-objective problem may in principle be two
different ways to pose the same underlying problem and can be solved by EAs (Karaboga,
2004). Evolutionary algorithms are proving robust in delivering global optimal solutions and
helping to resolve limitations encountered in traditional methods (Enitan and Adeyemo,
2011).
Like many other natural world problems, problems of AD process are conflicting in nature.
For instance, reduction of organic matter to meet the discharge standard and biogas
production during anaerobic treatment of wastewater requires different conflicting objectives
5
such as maximization of desirable properties (such as biogas production for energy
generation) and simultaneously minimizing its undesirable characteristics (such as a
reduction of effluent substrate concentration or the organic pollutant loads in the final
effluent to meet the discharge standards) (Babu et al., 2005). Evolutionary algorithms are of
interest in optimizing AD processes to generate Pareto-optimal solutions, although this may
not be an easy task due to the complexity and variations in the organic content of the
industrial wastewaters.
Recently, a few industries have started using sophisticated technologies to improve, monitor,
optimize and control processing parameters in order to increase treatment efficiency
(Rodríguez-Fernández et al., 2007; Iqbal and Guria, 2009). However, expert knowledge is
still needed to apply these techniques successfully. If the specific technique is not applicable
to certain problem due to unknown system parameters, then multiple local minima or non-
differentiability evolutionary algorithms have the potential to overcome these limitations
(Karaboga, 2004) by using mathematical model-based techniques to make decisions about
optimal production scenarios.
1.1 STUDY OBJECTIVES
1.1.1 Aim
The aim of this study was to monitor the performance and the microbial diversity especially
the methanogens in a UASB reactor treating brewery wastewater, develop a dynamic model
to describe the behaviour of a UASB reactor and optimize the model using an evolutionary
algorithm called the Combined Pareto Multi-objective Differential Evolution (CPMDE)
algorithm.
1.1.2 Objectives
In order to achieve our aim, the following specific objectives were set out:
To monitor the parameters associated with the performance of a full-scale UASB reactor
in order to establish kinetic constants to be used in further mathematical modelling.
6
To determine the microbial community structure of the full–scale UASB reactor using
different molecular techniques including; FISH, PCR and QPCR.
To develop and simulate a dynamic multi-variable optimization model in order to predict
the methane production rate during the anaerobic treatment of brewery wastewater using
MATLAB object-oriented language.
To optimize the developed (MMGM) and adopted modified Stover–Kincannon kinetic
models using a combined Pareto multi–objective differential evolution (CPMDE) algorithm
to maximize CH4 production and enhance wastewater treatment efficiency.
1.2 THESIS OUTLINE
Chapter one begins with a general introduction with the study objectives and outline of the
thesis. Each subsequent chapter is concluded with details of the research output(s) from the
chapter. Chapter two presents the literature review relevant to the study. Chapter three
presents the physico-chemical composition of brewery wastewater and the performance of
the full–scale UASB reactor treating the brewery wastewater in KwaZulu-Natal, South
Africa. Chapter four presents the identification and quantification of the microbial ecology of
the full–scale UASB reactor. The Stover–Kincannon kinetic model was adopted to predict
effluent substrate concentration, in order to reduce the pollutant load discharged into the
environment and water bodies. In chapter five, development of kinetic model (MMGM) for
CH4 production during AD was presented. Chapter six is a continuation of the study in
Chapter four and five where, a multi-objective constrained optimization problem was
presented. A novel evolutionary algorithm called CPMDE algorithm was used as the
optimization tool to integrate and determine the optimum CH4 production rate, effluent
substrate concentration and biomass washout from the UASB reactor treating brewery
wastewater. Finally, Chapter seven presents the general conclusions of the study along with
suggestions for future research and the significance of the study. The thesis ends with a list of
references and appendices.
7
CHAPTER TWO: LITERATURE REVIEW
2.1 INTRODUCTION
Manufacturing industries produce wastes that contain high levels of organic materials which
could adversely affect the environment should they be directly discharged. For industries to
meet discharge requirements, economical and practical treatment methods are important
factors that need to be considered. Therefore, there is an increasing need for industries to treat
their wastewater before discharged into the environment using effective, eco-friendly, simple
and inexpensive technologies, thereby minimizing the impact on the environment. Anaerobic
treatments have gained more attention in developing countries due to the fact that traditional
aerobic technologies, like the activated sludge process, require professional skills and high
costs to operate and may not be able to handle high strength effluents (Bhatti et al., 1993;
Leitao et al., 2006).
Anaerobic treatment involves the conversion of complex organic matter present in low to
high–strength industrial wastewaters into simpler monomers and production of biogas in a
closed system, through the activity of various anaerobic microorganisms (Bhatti et al., 1996;
Keyser, 2006; Ziemiński and Frąc, 2012; Enitan et al., 2014a). Biogas recovery systems
referred to as ‗methane (CH4) recovery systems‘, ‗bioreactor/biodigester‘, ‗methane digester‘,
or ‗anaerobic digester‘ can be used to treat industrial waste and capture CH4 that can be used
for on-site energy generation (Parawira et al., 2005). However, the reduction of organic
matter and quantity of biogas released depends on the conditions under which the reactor is
operated (Gyalpo et al., 2008), because any sudden changes in the performance of the system
can have a damaging effect on the quality of effluents discharged and biogas recovery.
Several anaerobic digestion (AD) technologies have been designed and constructed for the
treatment of high–strength wastewater (Demirel et al., 2010; Abbasi and Abbasi, 2012).
Anaerobic system such as an upflow anaerobic sludge blanket (UASB) reactors have received
much attention due to their ability to treat industrial wastewaters at higher organic loading
rate (OLR) and a lower hydraulic retention time (HRT) (Mata-Alvarez et al., 2000; Nadais et
al., 2011).
8
2.2 ANAEROBIC TREATMENT OF WASTEWATER
Anaerobic digestion has received worldwide attention due to it being a simple, inexpensive
technology to operate, that produces low biomass outputs and low energy input
(Karagiannidis and Perkoulidis, 2009; Kaparaju et al., 2010). The treatment of high-strength
industrial wastewater such as brewery wastewater using AD technologies has been employed
in several instances throughout the world (Brito et al., 2007; Demirel et al., 2010; Simate et
al., 2011). It has been used widely as a source of renewable energy. The biogas comprising of
carbon dioxide (CO2), CH4 and traces of other gases produced during the process of AD can
be used directly as fuel in combined heat and power gas engines, thereby reducing the release
of these biogases to the atmosphere (Ward et al., 2008; Singh and Prerna, 2009). On the other
hand, some disadvantages of AD processes includes long retention times (Chan et al., 2009),
bad odour and effluents that sometimes needing post-treatment to meet the discharge
standards for nutrients levels, organic matter and pathogens content (Seghezzo et al., 1998).
Over the past 25 years, different types of reactors have been developed and their installations
have been commercialized. Along with the UASB reactor (Fang et al., 1995a; Lettinga, 1995;
Ryan et al., 2010; Qiao et al., 2011; Chong et al., 2012), anaerobic sequencing batch reactor
(ASBR) (Shao et al., 2008; Won and Lau, 2011), hybrid upflow anaerobic sludge-filter bed
(UASFB) (Rajagopal et al. 2009), continuous stirred tank reactors (CSTR) (Diaz et al., 2006;
Klocke et al., 2007; Kaparaju et al., 2010; Mirzoyan et al., 2010), expanded granular sludge
bed (EGSB) (Seghezzo et al., 1998), anaerobic baffled reactor (ABR), anaerobic fixed-bed
reactors (AFBR) and membrane technology have been widely used for wastewater treatment
(Figure 2.1a) (Lettinga et al., 1980; Driessen and Vereijken, 2003; Parawira et al., 2005;
Zhou et al., 2006). Among these UASB reactor configuration is the most widely used high–
rate anaerobic reactor for the treatment of high-strength wastewater. Over one thousand
UASB reactors have been installed worldwide due to its simple design and low operational
cost (Tiwari et al., 2006; Nigel and Sneeringer, 2011). An overview of the above-mentioned
anaerobic treatment systems used for different industrial wastewater pre-treatment is
presented in Figure 2.1(b) (International Energy Agency, 2001). Descriptions and further
information on the different types of reactors can be found in the literature (Shao et al., 2008;
Sipma et al., 2010; Won and Lau, 2011; Rajagopal et al., 2013; Tauseef et al., 2013).
9
Figure 2.1: (a) Proportions and types of anaerobic digestion systems that have been installed
and commercialized for the treatment of industrial wastewater (b) percentage of industries
using anaerobic treatment technologies for industrial wastewater (International Energy
Agency, 2001).
2.2.1 Upflow Anaerobic Sludge Blanket Reactors
The UASB reactor designed by Lettinga et al. (1980) has made AD the most competitive and
favourable treatment technology for high–strength organic wastewaters (Ryan et al., 2010;
Abbasi and Abbasi, 2012). It has been widely employed to treat industrial and domestic
wastes around the world due to features such as simple design, easy construction and
maintenance, low operating cost, high removal efficiency, short retention time, stability,
temperature and low energy demand ( Alvarez et al., 2006; Tiwari et al., 2006). UASB
reactors are highly dependent on its granular sludge as the core component during wastewater
treatment for an effective conversion of organic matter to biogas (Batstone et al., 2002; Liu et
al., 2003).
Fluidised
bed
2% Hybrid
3% Lagoons
5%
Anaerobic
filters
7%
CSTR
8%
EGSB
8%
UASB
67%
a
Chemical
7% Pulp &
Paper
9%
Distillery
12%
Brewery &
Softdrink
25%
Food
40%
Various
7%
b
10
A schematic diagram of a typical UASB reactor is shown in Figure 2.2. In an UASB reactor,
the influent enter through the bottom of the reactor, thereby helping in the aggregation of
microbial biomass in the sludge bed and blanket to get in contact with the influent (Abbasi
and Abbasi, 2012). Several investigations have been carried out at laboratory, pilot and full-
scale level to optimize UASB reactors using different types of effluent including domestic
(Atashi et al., 2010) and industrial wastewaters. Some of the industrial effluents treated
include pharmaceutical (Herumurti et al., 2008), pulp and paper (Ali et al., 2009), sugar
factories (Demirel and Scherer, 2008a; Hampannavar and Shivayogimath, 2010) brewery
wastewater (Parawira et al., 2005; Kovacik et al., 2010; Madukasi and Zhang, 2010),
slaughterhouse (Nacheva et al., 2011) and textile (Muda et al., 2011).
Figure 2.2: Schematic diagram of an upflow anaerobic sludge bed (UASB) reactor with red
balls indicating granules and yellow balls indicating evolved biogas.
Influent
UASB effluent
Sludge bed
Biogas
Sludge blanket
Gas-liquid-solid seperator
Baffle
Gas
bubbles
Granules
11
2.3 BIOGAS RECOVERY FROM ANAEROBIC DIGESTERS
Due to the increasing effect of climate change in the world, industrial waste management
strategies and reduction of environmental effects caused by the industrial waste disposal has
gained more attention. From the clean development mechanisms (CDM) point of view,
‗mitigating CH4 emissions‘ is very fascinating, since the global warming potential (GWP) of
CH4 is 21 times higher than that of CO2. Under anaerobic conditions CH4, CO2, nitrogen
(N2), hydrogen (H2), hydrogen sulphide (H2S) and oxygen (O2) called ‗biogas‘ are produced
(Wen et al., 2009) with calorific values of 21-24 Mj/m3, equivalent to 6 KWh/m
3 of CH4
(Bond and Templeton, 2011). Moreover, the use of biomass for energy generation is
classified as a 'carbon neutral' process because, the CO2 released during this process is
balanced by the CO2 absorbed by plants during their growth (The Centre for Sustainable
Environmental Sanitation the Centre for Sustainable Environmental Sanitation, 2009).
Furthermore, the use of CH4 gas from AD as a renewable energy source has been widely
adopted as one of the CDM in order to obtain a certified emission reduction (CER) credit
under the Kyoto Protocol. This facilitates the promotion of biogas to reduce the greenhouse
effect, through reduction of CH4 emissions into the atmosphere. Biogas generation has been
widely adopted in Asia, particularly in Bangladesh, China, India and Nepal (Dutschke et al.,
2006; NSWAI-ENVIS, 2007; The Centre for Sustainable Environmental Sanitation, 2009).
The treatment of these wastes and biogas production with high CH4 content as energy
recovered is a good alternative to fossil fuel, since human activities and industries at large
produce sufficient amounts of waste (Tauseef et al., 2013). The biogas that is used for fuel
energy must contain more than 50% of CH4 (Srisertpol et al., 2010) which could be used for
heating, cooking, lighting or to generate electricity for domestic and larger industrial plants
(Bond and Templeton, 2011; Heffernan et al., 2012).
However, South Africa and other African countries have placed less attention on the
implementation of national biomass energy from wastewater when compared to world‘s
implementation of AD technology. The application of AD technology in South Africa for the
treatment of industrial wastewater has been reviewed by Ross and Louw, (1987) and Stafford
et al., (2013). Their survey showed that the use of anaerobic reactors, especially UASB
12
reactors that could be used for recovery of energy is very low in South Africa compared to
other countries.
In 2003, the South Africa Government set a ten-year target to produce 10 000 GWh (0.8
MTOE) from biomass, wind, solar and small-scale hydropower technology for renewable
energy consumption by 2013. The energy from biomass could be used for power generation
and non-electric technologies such as bio-fuels and solar water heating. It is approximately
1667 MW (4%) of the estimated electricity demand for 2013 (41539 MW)—which is
equivalent to replacing two units of Eskom's combined coal fired power stations (2 x 660
MW) (Shabangu, 2004).
In another report for biogas generation for electricity in 2012, environmental engineering
company Talbot & Talbot was employed to design and supply a biogas train in South Africa–
for its first green energy project (Cloete, 2008). The R5-million biogas project, include the
production of biogas from anaerobic wastewater digestion that can be converted to electricity
or used as boilers fuel (Cloete, 2008). However, some industries and municipal wastewater
treatment plants that have onsite anaerobic reactors still flare or vent the biogas that are
produced during anaerobic treatment of wastewater (Stafford et al., 2013). This demonstrates
that energy use has been poorly integrated and the opportunities for mitigating greenhouse
gas (GHG) emissions have not been realized.
Few industries in South Africa such as Cape Flats wastewater treatment plant in Cape Town,
PetroSA's gas-to-liquids refinery in Mossel Bay and some isolated community, household
and small-scale industries are using biogas generated during anaerobic treatment for energy
generation (Stafford et al., 2013). Stafford et al. (2013) further listed different types of
industrial and domestic blackwater wastewaters being treated using anaerobic reactors in
South Africa.
13
2.4 BIOCHEMISTRY AND MICROBIOLOGY OF THE ANAEROBIC
DIGESTION PROCESS
The AD process is carried out by a group of facultative, obligate and strict anaerobic bacteria
(Mshandete et al,. 2005; Appels et al., 2008) that are divided into four groups (Figure 2.3)
based on the biochemical processes and the metabolites they produce. Under ideal conditions,
these microorganisms break down the complex organic compounds through a variety of
intermediates into the components of biogas, such as CH4 and CO2 with small levels of H2S,
H2 and N2 (Appels et al., 2008; Mirzoyan et al., 2010; Amani et al. 2011). The overall
reaction is shown in equation 2.1 (Bitton, 1994).
Organic matter CH4 + CO2 + H2 + NH3 + H2S (2.1)
About 70% of the total CH4 production during AD is from acetic acid, while the remaining
30% comes from H2 and CO2 conversion (Ahring, 2003). It has been reported that about 80 -
90% CH4 composition can be produced in reactors treating wastewater (Okonkwo et al.,
2013). The origin of the AD process and the biodegradable materials determines the
composition of biogas produced.
The stability of the microbial ecosystem in the AD process has been shown to depend on the
methanogenic activity, which is characterized by slow growth rates of microorganisms. These
microorganisms have been found to be very sensitive to operational and environmental
variations in the anaerobic wastewater treatment systems, such as salinity, sludge properties,
temperature, pH, mineral composition, loading rate, HRT, carbon-to-nitrogen ratio and
volatile fatty acids (VFAs). These factors in-turn influence the digestibility of the organic
matter and production of biogas (Leitao et al., 2006; Chong et al., 2012).
14
Figure 2.3: The key stages of anaerobic digestion of organic matter in the wastewater (Li et
al., 2011).
2.4.1 Hydrolytic Bacteria
The digestion process is initiated by the action of facultative and obligate fermentative
anaerobic bacteria of mainly the genera Bifidobacteria, Lactobacillus, Enterobacterium and
Streptococcus (Krzysztof and Frac, 2012). This stage has been found to be common to both
aerobic and AD processes. The anaerobic bacteria were shown to catalyse the breakdown of
large complex soluble and insoluble organic molecules present in the wastewater into smaller
soluble monomers which could be transported into cells of non-hydrolytic fermentative
bacteria and metabolized (Bitton, 1994). The rate of hydrolysis process has been shown to be
Methanogenesis: Archaea
Homoacetogenesis
Reductive
Methanogenesis
(30%)
;
CH4, CO2
Aceticlastic
Methanogenesis
(40-70%)
Volatile Fatty Acids: Propionate, Butyrate,
etc.
Fermentation and Acidogenesis bacteria
Amino acids and simple sugars Long chain fatty acids
Hydrolysis
Lipids
Complex Biodegradable particulates
Proteins and carbohydrates
Acetogenesis
Acetic acid (CH3COOH) H2, CO2
15
dependent on parameters such as: wastewater type, pH, size of particles, production of
enzymes, diffusion and adsorption of enzymes on the particles of wastes subjected to the
digestion process (Ziemiński and Frąc, 2012).
2.4.2 Fermentative Acidogenic Bacteria
The acidogenic bacteria are the largest trophic groups, and consist of about 90% of the total
bioreactor population (Zeikus, 1980). Several microbial genera take part in acidogenesis, the
first stage of AD (Krzysztof and Frac, 2012). Acidogenesis is the process during which more
―simple‖ organic material is metabolized to form CO2, H2, acids and alcohols through the
action of the genera viz, Pseudomonas, Bacillus, Clostridium, Micrococcus or
Flavobacterium (Krysztof and Frac, 2012). This process has been divided into two stages:
The hydrogenation and dehydrogenation. The acids forming bacteria convert sugars, fatty
acids and amino acids to organic acids (including formic, acetic, propionic, butyric, lactic
acids), ketones and alcohols, which causes an accumulation of electrons in response to an
increase in H2 concentration in the solution. The new products may not be used directly by
methanogenic bacteria and must be converted by obligate anaerobes producing H2 in the
process called acetogenesis (Krzysztof and Frac, 2012). Other acid-forming bacteria facilitate
the production of acetate, H2 and CO2 depending on environmental conditions such as pH,
temperature for direct assimilation of the new metabolites as substrates and energy source by
the methanogens (Keyser, 2006; Krzysztof and Frac, 2012).
2.4.3 Acetogenic Bacteria
Acidogenesis products are further oxidized to acetate, H2, and CO2 by the activity of
acetogenic bacteria (acetate and H2-producing bacteria) such as Desulfovibrio,
Syntrophobacter wolinii, Syntrophomonas wolfei, Syntrophus buswellii, Syntrophococcus,
Natroniella and Acetigena spp. (Bhatti et al., 1996; Pitryuk and Pusheva, 2001; Karnholz et
al., 2002). The acetogens have been found to be obligate H2-producing bacteria that can only
survive at very low H2 concentrations. They have been shown to help in the conversion of
fatty acids and alcohols to acetate, H2, and CO2 at low H2 partial pressure (Bitton, 1994).
Therefore, for acetogenic bacteria to maintain a low partial pressure of H2 less than 10-5
atmospheres, they live in symbiosis with the H2-utilizing methanogens when the digester is
16
operated at optimum temperature and pH levels (Ziemiński and Frąc, 2012). Most digesters
are normally operated at about pH 7 because; it favours all the groups that are involved in the
conversion of organic matters to biogas including methanogens. Studies on the syntrophic
reactions have been described in the literature with optimum pH levels and temperature
between 6.3-8.5 at 25˚C and 45˚C respectively for syntrophic association of acetogenic and
methanogenic bacteria (Schink, 2002; Amani et al., 2011).
2.4.4 Methanogenic Archaea and their Taxonomy
Living organisms have been classified into three main taxonomies based on 18S and 16S
rRNA analysis and comparison of conservative phylogenetic features. The phylogenetic
domains include Archaea, Bacteria and Eukarya. Organisms belonging to domain Archaea
are divided into two phyla namely Crenarchaeota and Euryarchaeota (Figure 2.4) (Anderson
et al., 2009). The Crenarchaeota have been discovered to consist mainly of thermoacidophiles
and thermophiles while the Euryarchaeota contains a wider variety of organisms including
the methanogens, the extreme halophiles, thermoacidophiles and thermophiles. Recently the
third phylum, Thaumarchaeota, was proposed to include the mesophilic organisms previously
classified as Crenarchaeota (Brochier-Armanet et al., 2008). The CH4-producing organisms
(methanogens) are classified to domain Archaea, and phylum Euryarchaeota based on the
phenotypic and taxonomic classification (Ziemiński and Frąc, 2012).
Methanogenic bacteria are divided into four classes, five orders, nine families and 26 genera.
They are different from each other in shape, membrane lipids, 16S rRNA sequence, structure,
cell wall chemistry and other features (Demirel and Scherer, 2008b, Ziemiński and Frąc,
2012). Figure 2.4 shows the phylogenetic classification of methanogens. Methanogens are
archeons, unlike bacteria, they do not have a typical peptidoglycan (mureinic) skeleton; rather
several genera have pseudomurein, while others have walls consisting of lipids composed of
isoprenoid hydrocarbons glycerol lipids with different metabolism (Ziemiński and Frąc,
2012). Methanogenic ribosomes exhibit a similar size to that of eubacteria ribosome, but their
sequence of ribosomal RNA is completely different (Watanabe et al., 2004).
17
Figure 2.4: Classification of methanogens based on 18S and 16S rRNA analysis and comparison of conservative phylogenetic features (Demirel
and Scherer, 2008b; Ziemiński and Frąc, 2012).
Methanothermo
bacter
Methanobacteriaceae
Methanobacterium
Methanospaera
Methanobrevibacter
Methanothermoceae
Methanothermus
Methanococcaceae
Methanococcus
Methanocaldococcaceae
Methanocaldoccocus
Methanothermo
coccus Methanotorris
Nanoarchaeota
Methanopyrales
Methanopyri
Methanopyrus
Methanopyraceae
Methano
sprillaceae
Methanomicrobiales
Methanomicrobia
Methanosarcinales
Methanoculleus
Methanolacinia
Methanomicrobium
Methanogenium
Methanofollis
Methanocorpu
sculum
Methanocorpu
sculaceae
Methano
microbiaceae
Methanoplanus
Methano
spirillum
Methanomicrococcus
Methanohalophilus
Methanosalsum
Methanomethylovorans
Methanohalobium
Methanococcoides
Methanosarcina Methanosaeta
Methano
sarcinaceae
Methano
saetaceae
Methanolobus
Archaea
Euryarcheota Crenarcheota Korarcheota
Methanobacteria
Methanobacteriales
Methanococci
Methanococcales
18
Methanogens are largely differentiated morphologically. They exhibit almost all shapes
occurring in bacteria including cocci (Methanococcus), rods (Methanobacterium), short rods
(Methanobrevibacter), spirillaceae (Methanospirillum), sarcina (Methanosarcina) and
filiforms (Methanothrix). The size of these microorganisms ranges from 0.3 to 7.4 μm
(Karakashev et al., 2006). They are strict anaerobes and contain neither catalase nor
superoxide dismutase. Due to extraordinary sensitivity of methanogens to oxygen, their
biochemistry, physiology and ecology have been reviewed (Ziemiński and Frąc, 2012). Some
of their characteristics include, their sensitivity to changes in pH and temperature, inhibition
of their growth by high level of H2, sulphur, NH3 and VFAs and other compounds, in the
environment or in the bioreactor (Ziemiński and Frąc, 2012, Nakasaki et al., 2013).
Methanogens are slow-growing bacteria with a generation time between 3 days at 35˚C and
50 days at 10˚C (Bitton, 1994). Studies have shown that three different major pathways exist
for CH4 formation depending on the source of the reducing potential and the carbon
compound used as substrate (Figure 2.5); which include hydrogenotrophilic, aceticlastic and
the methylotrophic methanogens (Bapteste et al., 2005; Ziemiński and Frąc, 2012).
Hydrogenotrophilic methanogens are H2 using organism. They use H2 as an electron donor to
reduce CO2 to CH4 (Figure 2.5; Equation 2.2). This group helps in maintaining very low
levels of partial pressure needed by the aceticlastic methanogens for the conversion of VFA
and alcohols to acetate (Gerardi, 2003).
CO2 + 4H2 CH4 + 2H2O (2.2)
Abundance of Methanobacterium, Methanobrevibacter and Methanococcus of orders
Methanobacteriales, Methanomicrobiales and Methanococcales in different types of
anaerobic bioreactor treating wastewaters have been reported (Casserly and Erijman, 2003;
Bhatti et al., 1993; Liu et al., 2002a; Castro et al., 2004; Diaz et al., 2006; Cardinali-Rezende
et al., 2009; Kovacik et al., 2010a). The second group is the acetotrophic or aceticlastic
methanogens which convert acetate to CH4 and CO2 (Zheng and Raskin, 2000). The overall
reaction is;
CH3COOH CH4 + CO2 (2.3)
19
Figure 2.5: Pathways of methanogenesis: hydrogenotrophic (double-lined arrows),
aceticlastic (solid arrows) and methylotrophic (broken gray arrows) (Bapteste et al., 2005).
In the aceticlastic pathway, the CO2 is oxidized to provide electrons and the methyl group
converted from acetate is linked to methanopterin (or sarcinapterin, for Methanosarcina)
before being reduced to CH4 in two enzymatic reactions (Figure 2.5), the last two steps of the
hydrogenotrophic pathway (Meuer et al., 2002). The most commonly reported aceticlastic
methanogens from bioreactors belong to the genera Methanosaeta (coccoid bacteria) and
Methanosarcina–sheathed rods or long filaments bacteria (Keyser et al., 2006). This group of
methanogens helps in the production of about 70% of the total CH4 generated during the AD
of wastewater (Ahring, 2003). Methanosaeta sp. such as M. thermophila and M. concilii
belonging to genus Methanosaeta utilize acetate, while Methanosarcina strains like M.
barkeri, M. mazeii and M. thermophila utilize acetate, methanol, methylamines, H2 and CO2
as substrate (Keyser et al., 2006). The abundance of Methanosarcina sp. at high acetate levels
and Methanosaeta sp. at low acetate concentrations has also been reported (Keyser et al.,
20
2006). An abundance of Methanosarcina and Methanosaeta sp. in UASB granules treating
different wastewaters at steady–state conditions have been reported in the literature (Fang et
al., 1994, 1995b; Chan et al., 2001; McHugh et al., 2003b).
The third group is the methylotrophic methanogens. This group include orders
Methanosarcinales and Methanobacteriales namely Methanosarcina barkeri and genus
Methanosphaera (with several possible variants) respectively (van der Wijngaard et al.,
1991; Meuer et al., 2002). They directly produce CH4 from methyl groups (-CH3),
methylamines [(CH3)3-N] and methanol (CH3OH) as substrate (Gerardi, 2003). Methanol is
usually found as organic pollutant in several wastewaters and is a substrate for both
methanogens and acetogens (Bapteste et al., 2005). Compounds such as methanol or methyl-
amines can be used as both electron acceptor and donor respectively. In running the
hydrogenotrophic pathway in the reverse direction from methyl-CoM to CO2, one molecule of
C-1 compound is oxidized to provide electrons for reducing three additional molecules to
CH4. However, some Methanosarcinales can reduce this C-1 compound in the presence of
methanol and H2/CO2, using only the last step of hydrogenotrophic methanogenesis (methyl-
CoM to CH4), drawing electrons from H2 (Bapteste et al., 2005).
2.4.5 Techniques Used To Detect Microorganisms from Anaerobic Reactor Samples
The main aims of studying any microbial ecology include the identification, classification
and determination of microbial activity in the granules of an anaerobic reactor (Ziemiński and
Frąc, 2012). In the past, traditional identification methods have been used to determine the
morphology and phenotypic characteristics (Smith, 1966; Zeikus, 1977; Grothenhuis et al.,
1991; Liu and Tay, 2002), which are time-consuming and limited. Many microorganisms
especially the methanogens are difficult to culture using traditional methods, because they
have slow growth rates, restricted environmental conditions and selective nutritional
requirements (Grothenhuis et al., 1991; Briones and Raskin, 2003).
The development of molecular techniques (Figure 2.6) to study the complex microbial
populations in environmental samples has eliminated the use of more elaborate traditional
21
techniques of culturing microorganisms (Gonzalez et al., 2003; Hofman-Bang et al., 2003).
Basically these techniques have been divided into two main types: quantitative and
qualitative. Qualitative techniques that may be used include polymerase chain reaction based
denaturing gradient gel electrophoresis (PCR-DGGE), temperature gradient gel
electrophoresis (TGGE) and terminal-restriction fragment length polymorphism (T-RFLP)
etc. Microbial profiling techniques involve amplifying the nucleic acids isolated from
environmental samples, sequencing and comparing them to the known sequences in the
GenBank database appropriate for identifying related microorganisms. These methods have
been successfully employed to study complex microbial populations in the laboratory- and
industrial-scale fermenters to study the shift in microbial community structure (McHugh et
al., 2003a; Wang et al., 2010; Ziganshin et al., 2011; Shen et al., 2013).
Quantitative real-time PCR (QPCR) and fluorescence in-situ hybridization (FISH) are
quantitative techniques used in the survey of microbial ecologies (Yu et al., 2006; Zhang and
Fang, 2006; Demirel and Scherer, 2008b; Tabatabaei et al., 2009; Bergmann, 2012; Traversi
et al., 2012). Fluorescence in-situ hybridization has also been used for the quantitative
analysis and to understand the spatial distributions of microorganisms (Briones and Raskin,
2003). This technique is based on hybridization of whole cells with specific probes, and
microscopic analysis of dyed hybridized cells using epifluorescence microscopy, flow
cytometry or scanning electron microscopy (Demirel and Scherer, 2008b; Tabatabaei et al.,
2009).
Quantitative real-time PCR on the other hand, can be used to amplify and simultaneously
quantify targeted DNA sequence by employing a PCR-based technique that enables one to
quantify the number of gene copies or relative number of gene copies in a given sample. The
reliability of QPCR results is strongly dependent on the quality of the extracted genomic
DNA (Bergmann, 2012). The amplified gene copy number from bulk DNA reflects the
relative abundance of the microorganisms in the community. The amplification principle of
QPCR is similar to that of PCR. This technique monitors the concentration of the amplified
target after each PCR cycle using a fluorescent dye or probe change in fluorescence intensity
that reflects the concentration of amplified gene in real-time assay (Zhang and Fang, 2006;
Bergmann, 2012). Either absolute or relative quantification can be used to determine the
22
concentration of DNA or RNA in an extracted sample. This technique has been widely used
to quantify the microbial population and dynamics in anaerobic reactors in their natural
environments (Yu et al., 2005; Yu et al., 2006; Traversi et al., 2012).
However, it is difficult to monitor specific groups or a domain using only one technique as
each technique has its own merits and demerits. Therefore, a combination of qualitative and
quantitative methods including PCR-DGGE, QPCR and microarrays could be used to
overcome the limitations of one technique (Park et al., 2009). A combination of different
molecular techniques, such as electron microscopy, PCR-based DGGE, cloning and FISH to
gain insight into the physical appearance, function and structure of microbial diversity of
methanogenic granules from a full-scale UASB reactor treating brewery wastewater have
been explored in the past (Diaz et al., 2006). The PCR-based DGGE and FISH analyses were
used to identify the microbial populations in a full-scale UASB reactor treating brewery
wastewater (Chan et al., 2001; Liu et al., 2002a). Chan et al. (2001) reported Delta and
Gammaproteobacteria; Methanosaeta concilii and Methanobacterium formicicum as the
dominant bacterial and Archaea bands detected in the full-scale UASB reactor. Likewise,
Keyser et al. (2006) used PCR-DGGE for the fingerprinting and identification of the
microbial consortium present in different types of granules collected from the UASB reactor
treating brewery wastewater. Diverse group of methanogens such as Methanosarcina,
Methanosaeta, Methanobacterium and uncultured bacteria belonging to Archaea domain
were identified and fingerprinted using PCR-DGGE technique (Keyser et al., 2006).
23
Figure 2.6: Flow diagram of different steps used in studying the structure and functions of
microbial communities in an environmental sample.
Isolation
Ad
ap
tati
on
of
cult
ure
co
nd
itio
ns
RT-PCR QPCR
Nucleic acid extraction
PCR
PCR Products
DGGE OR T-RFLP
Cloning
Nucleic acid extraction
QPCR
Sequencing of unique
clones
Sequencing of
unique clones
Phylogenetic Tree Primers and Probes
Hyb
rid
izati
on
an
aly
sis
Genetic Fingerprint
using FISH
Genetic
Fingerprint
s
Clone
Libraries Quantification
Comparative sequencing analysis
PCR
Cultures
DNA
Environmental Sample
RNA
Sequencing Database
Nucleic
acid
extractio
n
24
2.5 FACTORS AFFECTING PERFORMANCE OF UASB REACTORS AND
BIOGAS PRODUCTION
Even though the advantages of using anaerobic systems for pre-treatment of wastewater are
recognized, anaerobic treatment plants are subjected to variations in one or more parameters,
which affect or define the reactor‘s performance (Leitao, 2004; Keyser, 2006). Concerns still
exist about reactor stability, effluent variability, the biological degradation of the adsorbed
organic matter and activities. Due to these facts, several works on the operational variations
and the stability of UASB reactor performance due to extreme transient conditions have been
reported (Turovskiy and Mathai, 2006; Coelho et al., 2007; Abbasi and Abbasi, 2012). The
operational pH, temperature, nutrients availability, presence of VFA, influent COD
concentration, influent type, sludge retention time (SRT), organic and hydraulic load
variations are some of the parameters that should be monitored for the successful operation of
any anaerobic reactor treating industrial wastewater (Fang et al., 1995a; Leitao et al., 2006;
Coelho et al., 2007; Abbasi and Abbasi, 2012).
2.5.1 Organic Loading Rate
Organic loading rate (OLR) is an important parameter that significantly affects the microflora
and the performance of a UASB reactor. Fluctuations in organic load depends on the SRT,
HRT, sludge properties, mixing intensity, duration of the variation, bacterial mass and
activity (Rincón et al., 2008; Abbasi and Abbasi, 2012). Different studies have shown that
higher values of OLR can cause reduction in COD removal efficiency in a wastewater
treatment system (Torkian et al., 2003; Sánchez et al., 2005). Zhou et al. (2007) have
reported that a higher loading rate could cause unrecoverable acidification, suppression of the
methanogenic activity due to serious imbalance between the methanogens and the acidogens,
as well as inhibition of methanogens by VFA production (Latif et al., 2011).
2.5.2 Nutrients
The ability of anaerobic microorganisms to grow depends on the availability of the essential
nutrients that are present in the wastewater (Speece et al., 1983; Lettinga, 1995; Fermoso et
al., 2008; Mudhoo and Kumar, 2013). Lack of these nutrients could negatively affect their
growth and the efficiency of the anaerobic degradation (Lettinga, 1995; Mudhoo and Kumar,
25
2013). The biochemistry of fermentation and CH4 production involves many enzymes that
contain different trace elements that need to be supplied as nutrients. Each anaerobic
microorganisms involved in the degradation of complex organic matter to simple components
are trace element specific, depending on the enzyme pathways (Zandvoort et al., 2006).
Several studies on the impact of nutrients on the efficiency of AD and enhancement of
granules in the bioreactors have been reported (Yu et al., 2001; Zandvoort et al., 2006;
Fermoso et al., 2008; Krishna, 2013; Zhang et al., 2013). Some bacteria, such as CH4-
forming bacteria in the reactors have relatively high internal concentrations of iron, cobalt
and nickel (Zandvoort et al., 2006; Zhang et al., 2013), which may not be present in
sufficient concentrations in the wastewater produced from the industries. Therefore, the
addition of trace elements prior to treatment to improved reactors performance is highly
recommended (Yenigun et al., 1996; Onodera et al., 2013). The optimum C: N: P ratio to
enhanced CH4 yield was reported to be 100:2.5:0.5 (Rajeshwari et al., 2000). This could be
calculated based on the wastewater biodegradable COD concentration, nutrient concentration
in bacterial cells and cell yield (Hulshoff-Pol, 1995).
2.5.3 Hydraulic Retention Time
The hydraulic retention time (HRT) has been defined as the average time that wastewater
spends inside the reactor (Bitton, 1994). The flow rate and composition of wastewater
entering the UASB reactor both affect the HRT (Cheng and Liu, 2002). High HRT increases
the contact time of wastewater with the sludge, thus improving the effluent quality and biogas
production rate. Therefore, a suitable HRT is very important for proper wastewater treatment
in a UASB reactor for better treatment efficiency as well as quality and quantity of biogas
concentration. From equation (2.5), the flow rate is inversely proportional to the HRT and
directly related to the reactor volume (Liu et al., 2003). This shows that volume is an
important parameter that must be considered when designing a reactor.
26
(2.5)
Where Q = Flow rate of influent stream (L/ d),
V = Volume of the reactor (L),
HRT = Hydraulic retention time (days).
Several studies have shown the effect of HRT on the microbial degradation in a single UASB
reactor treating different types of industrial wastewater (Diamantis and Alexandros, 2007;
Krakat et al., 2010; Muda et al., 2011).
2.5.4 Volatile Fatty Acids
Volatile fatty acids (VFAs) are important intermediate products in the formation of CH4. The
right concentration determines the efficiency of substrate removal from the reactor. For a
typical reactor, overloading, or sudden variations in HRT and OLR could cause the
accumulation of VFA and stressful conditions in the reactor during the break down of
complex organic matter (Wang et al., 2009). It can also affect the type of intermediates
produced. This might cause a shift between acetogens and acidogens population (VFA
producers), nitrogen reducing bacteria (NRB), sulphate reducing bacteria (SRB) and
methanogens (consumers) leading to drastic changes in biogas production rates and
compositions (Inanc et al., 1999; Akarsubasi et al. 2006; Wang et al., 2009). The toxic
effects of all VFAs in the AD process especially propionate, on the activity of acetogens and
methanogens have been investigated (Gallert and Winter, 2008; Uneo and Tatara, 2008;
Wang et al., 2009). Therefore, VFAs should be monitored and parameters adjusted in order to
avoid their accumulation in the UASB reactor to prevent the inhibition of methanogenic
organisms, thus reducing biogas production.
2.5.5 Operational Temperature
Operational temperature is an important parameter in anaerobic degradation processes. It
determines the dominant bacterial flora and the growth rates of microorganisms present in a
reactor (Khemkhao et al., 2012). Different species of bacteria can survive at different
27
temperature ranges. Operational temperature greatly affects the biodegradation and the biogas
production rate of any anaerobic reactor (Singh and Viraraghavan, 2000). The temperatures at
which anaerobic reactors could operate include psychrophilic (0-25°C), mesophilic (25-40°C)
and thermophilic conditions (40-60°C) (Sánchez et al., 2006). Studies have shown that the
performance of anaerobic reactors such as UASB reactors is temperature dependent (Kim et
al., 2006; Chen et al., 2008; Thomas, 2010; Khemkhao et al., 2012).
2.5.6 Operational pH
Microbial communities in the anaerobic digester have been shown to be highly pH dependent
and require suitable conditions of pH to grow optimally (Table 2.1). pH far below or higher
than the range required by the anaerobes could cause an accumulation of acetate, thereby
inhibiting the methanogens and leading to conversion of COD to volatile acids instead of CH4
production (Thomas, 2010). Therefore, most large scale AD reactors have been operated at
pHs of between 6.5–7.5. The standard operating method to keep the pH in this range has
been found to be the addition of lime and bicarbonate salts (Droste, 1997; Gerardi, 2003) or
by reusing treated effluent in the reactor (Najafpour et al., 2006; Espinoza-Escalante et al.,
2009). Therefore, controlling the pH of bioreactor is an essential factor for the growth of
diverse group of microorganisms and high reactor performance.
Table 2.1: Optimum pH ranges for selected methanogens
(Gerardi, 2003; Steinhaus et al., 2007)
Genus pH range
Methanothermus
Methanohalobium
Methanolacinia
Methanomicrobium
Methanosphaera
Methanogenium
Methanosprillium
Methanosaeta
Methanolobus
Methanothrix
Methanococcoides
6.5
6.5 – 6.8
6.6 – 7.2
7.0 – 7.5
6.8
7.0
7.0 – 7.5
7.6
6.5 – 6.8
7.1 – 7.8
6.5 – 7.5
28
2.6 MODELLING OF ANAEROBIC DIGESTION SYSTEMS
Mathematical models are analytical abstractions of the real world, representing the real
system and can be used to simulate the behavior of any system under investigation. Models
are typically a computer program, a set of mathematical formulas or an existing idea. Process
modelling is the design and description of a real system that provides a better understanding
of the processes, functions and its optimal working conditions (Pontes and Pinto, 2006). It
can also be used to control a process, predict a system‘s behavior and outcomes; without a
model, good predictions become difficult to make. Thus, among many other important
monitoring and control strategies for proper understanding of the underlying phenomena in
AD and biogas production is the development of suitable models, which adequately describe
processes taking place in the AD bioreactor. It is an elegant and cost-effective tool to
investigate certain engineering questions without wasting time and performing expensive
laboratory tests (Thorin et al., 2012). This has been found to be a good tool for process-
control strategies and to enhance gas production. Mechanistic and empirical or data-based
models are the two basic types of models available.
Mechanistic models are based on the underlying chemistry and physics governing the
behaviour of a process. They have a structure that clearly represents the biological, chemical
or the physical laws to propose one or more possible alternatives (Barampouti et al., 2005). A
mechanistic model assumes that a complex and real system can be understood by examining
the working and manner in which individual parts are coupled (Batstone et al., 2002).
Procedures for developing a mechanistic models include; the use of fundamental knowledge
of the interactions between process variables to define the model structure, the determination
of model parameters using experimental data, collection of data from the process to validate
the model; then if the model is not satisfactory, one can re-examine process knowledge and
restructure the model (Batstone et al., 2002; Mu et al., 2008; Yetilmezsoy, 2012)..
Based on qualitative understanding of UASB process gained over the years, several attempts
have been made to develop mechanistic models for quantitative descriptions of UASB
reactors (Colussi et al., 2012; Barampouti et al., 2005; Zhao et al., 2010; Elnekave et al.,
2012; Yetilmezsoy, 2012). A comprehensive model of AD processes known as anaerobic
29
digestion model no. 1 (ADM1) was proposed (Batstone et al., 2002). This model divided the
reactions that take place in the digester into two main types, biochemical and physico-
chemical reactions. Detailed description of the model can be found elsewhere (Batstone et al.,
2002). This model has been widely used in anaerobic processes for CH4 production (Mu et
al., 2008; Derbal et al., 2009; Zhou et al., 2011).
Wu et al. (2005) applied the axial dispersion model developed by Gomes et al. (1998) to a
laboratory scale UASB reactor using an orthogonal collocation algorithm. However,
mechanistic models have been found to be insufficient to understand the UASB process due
to several shortcomings in the models and to predict biogas production rates. These include
inaccurate prediction of substrate availability to the methanogens, or the rate of VFA
production or composition in the reactor (Elnekave et al., 2012). Other deficiencies in
formulation due to insufficient qualitative understanding of the process dynamics in reactor
have been reviewed in the literature (Sinha et al., 2002). These may be overcome through the
empirical observation and analysis of experimental data on UASB reactor performance
(Elnekave et al., 2012).
Empirical models are based on direct observation, measurement and extensive data records.
They are frequently used as the basis for process control designs. Response surface
methodology (RSM), fuzzy models and most recently, neural networks have emerged as one
of the most efficient methods in empirical modeling, particularly for non-linear systems (Abu
Qdais et al., 2010; Thorin et al., 2012). These models have been used to explain and predict
the performances of UASB reactors treating different wastewater from domestic and
industrial sources (Tay and Zhang, 1999; Holubar et al., 2002; Cakmakci, 2007).
Empirical modeling depends on the availability of representative data for model-building and
validation. Knowledge about the process is not needed for empirical modelling apart from
cause-and-effect between variables; empirical modelling uses a trial and error approach
(Thorin et al., 2012). This type of model does not require much data. Once the structure of
the model is defined, numerical techniques can be applied to parameterize the model. In this
case, although the structure has been determined from process knowledge, the modelling
30
procedure becomes an empirical one. The numerical techniques that are used are also very
different from those usually encountered in purely empirical modelling. They tend to be
iterative, and are more complex (Khataee and Kasiri, 2011).
Kanat and Saral (2009) developed an artificial neural network (ANN) model to study biogas
production from a thermophilic digester based on OLR, influent and effluent total VFA,
alkalinity, pH and temperature of the reactor. A similar study was also conducted by Abu
Qdais et al. (2010), where an ANN based model was developed to optimize CH4 production
using total and volatile solids, pH and temperature. Other studies on modeling of biological
and wastewater treatment processes using an ANN was reviewed by Khataee and Kasiri
(2011). The authors concluded that ANN models could predict the behaviour of the processes
based on experimental data with high correlation coefficients. They further mentioned that
additional information about the mechanisms and kinetics of the biological reactions was not
necessary. However, none of the studies reviewed by Khataee and Kasiri (2011) had biogas
production as an output parameter for their models. Ericson et al. (2010) modeled biogas
production from a full-scale biogas digester using process data obtained from several years of
running the digester using a statistically based ANN models.
Other model-based approaches to predict biogas production in an AD process have been
reviewed in the literature (Lyberatos and Skiadas, 1999; Levstek and Lakota, 2012; Thorin et
al., 2012). Regression neural network (GRNN), feed-forward back (FFBP) and radial basis
function-based neural networks (RBF) were designed and trained to predict the effluent
COD, TSS, and biogas production from a full-scale UASB reactor treating juice wastewater
(Elnekave et al., 2012). The ANN results reported for the prediction of both COD
concentration and biogas production were more accurate and closely related to the actual
biogas produced, while relatively larger discrepancies existed for the TSS concentration
(Elnekave et al., 2012).
31
2.7 OPTIMIZATION TECHNIQUES USING EVOLUTIONARY ALGORITHMS
In recent times, due to problems in evaluating the first derivatives, to locate the optimal for
many rough and discontinuous optimization surfaces, several free derivative–optimization
algorithms have been developed. This optimization problem is represented as an intelligent
search problem, where one or more agents are used to represent the constrained surface and
finding the optimal point on the search landscape (Das et al., 2008; Adeyemo and Otieno,
2009a). This includes restricting some variables of the system to be within certain ranges.
Evolutionary algorithms (EAs) as computer-based, biologically-inspired optimization
algorithms are stochastic searching methods commonly used for solving non-differentiable,
non-continuous and multi-modal optimization problems based on Darwin‘s natural selection
principle (Enitan and Adeyemo, 2011; Thorin et al., 2012; Sendrescu, 2013). They imitate the
process of natural evolution and are becoming more important optimization tools for several
real world applications for finding global optimum solutions regardless of initial parameter
value (Kachitvichyanukul, 2012.). General steps for evolutionary algorithm development are
shown in Figure 2.7. Evolutionary algorithms operate on a population of potential solutions,
applying the principle of survival of the fittest method to produce successful and better
solutions using evolutionary resembling operations (selection, reproduction and mutation),
which are applied to individuals in a population (Ronen et al., 2002; Shaheen et al., 2009).
The use of EAs in conjunction with a simulated model for an optimization is an important
factor for efficient and stable biogas production, especially CH4 (Adeyemo and Enitan, 2011;
Sendrescu, 2013).
32
Figure 2.7: Flowchart for evolutionary algorithm development.
Genetic algorithms are computerized search and optimization heuristics based on the
mechanics of natural genetics and selection (Deshmukh and Moorthy, 2010). They mimic the
natural evolution to make a search process. The solutions are commonly represented as
strings of binary digits. These algorithms require long processing times for a near-optimum
solution to evolve. These algorithm types have been successfully used in science and
engineering applications to reach near-optimum solutions to a variety of problems since its
introduction by Holland (1975). Details on the principal steps of a typical GA have been
reviewed (Mukhopadhyay et al., 2009; Enitan and Adeyemo, 2011). New GA techniques that
use real numbers for coding and genetic operators to generate new solutions until a stopping
criterion is satisfied have emerged (Mohebbi et al., 2008). It involves repeated procedures
with an initial population of potential solutions, a fitness evaluation via the application of
genetic operators and the development of a new population (Goñi et al., 2008). There are
different improved versions of the original genetic algorithm that have been reported in the
literature such as elitist non-dominated sorting genetic algorithm (NSGA-II) (Deb et al.,
2000), compressed-objective genetic algorithm (COGA-II) (Boonlong, 2013) and multi-
objective uniform-diversity genetic algorithm (MUGA) (Nariman-Zadeh et al., 2010).
No
Start
Generate initial population
Evaluate fitness values
Time to stop (Fitness value = Preset criteria) Stop
Generate new population
Yes
33
In an attempt to reduce the processing time and improve the quality of solutions, a differential
evolution (DE) strategy was introduced by Storn and Price (1995) for faster optimization.
Differential evolution is a population based algorithm like genetic algorithms using similar
operators; crossover, mutation and selection of optimization problems. The basic steps of a
DE algorithm are summarized in Figure 2.8. Differential evolution generates a new solution
by combining several solutions with the candidate solution. The population of solutions in
DE evolves through repeated cycles of three main DE operators: mutation, crossover, and
selection. Details on the DE operators and operation of the DE algorithm are discussed by
Deng et al. (2013) and Huang and Chen (2013). Unlike conventional GA that uses a binary
coding for representing problem parameters, the DE algorithm represents each variable in the
chromosome by a real number. Differential evolution selection process and its mutation
scheme make DE self–adaptive. Differential evolution has efficient straight forward features
that make it very attractive for numerical optimization. The basic approach of differential
evolution algorithm works as follows:
1. Initialize the number of a potential population (NP) at random, the maximum numbers
of evolution, the crossover rate (CR) and the scale factor, (F).
2. Initialize the population (pop), by some repair rules such that ‗variables‘ values are
within their boundaries.
3. Following the DE/rand/1/bin strategy, and production of new generation of individual
solutions:
a) Implementation of differential strategy on the individual mutation for each target
vector. The mutation component is a different vector of the parent.
b) With the crossover probability, each variable in the main parent is perturbed by
adding to it a ratio F of the difference between the two values of this variable in the
other two supporting parents. At least one variable must be changed. This process
represents the crossover operator in DE.
c) Selection operation of best solutions, if the resultant vector is better than the trial
solution, it replaces it; otherwise the trial solution is retained in the population.
d) Go to 2 above.
4. If the termination conditions are met go to 5, else go to 2 above
5. End.
34
Figure 2.8: Flowchart for the main steps in DE algorithm development.
The principal difference between DE and GA is that GA relies on crossover, a mechanism of
probabilistic and useful exchange of information among solutions to locate better solutions,
while DE uses mutation as the primary search mechanism (Enitan and Adeyemo, 2011).
However, the operators are not all exactly the same as those with the same names in GA
(Kachitvichyanukul, 2012). Differential evolution uses non-uniform crossover and
tournament selection operators to create new solution strings. In GA, two parents are selected
for crossover and the child is a recombination of the parents. In DE, three parents are selected
for crossover and the child is a perturbation of one of them. The new child in DE replaces a
randomly selected vector from the population only if it is better than it. In simple GA,
children replace the parents with some probability regardless of their fitness. All solutions in
DE have the same chance of being selected as parents without depending on their fitness
value. Differential evolution employs a greedy and less stochastic approach to solve problems
than the classical evolutionary algorithms (Babu and Chaurasia, 2003; Karaboga, 2004;
Selection of better vector
between target vector and
trail vector
Generate initial population and
fitness evaluation
No
Start
Termination
Mutation
End
Yes
Recombination
Fitness evaluation of trail
vector
35
Mariani et al., 2008; Liu and Wang, 2010). The crucial idea behind DE is a scheme to
generate the trial parameter vectors that are completely self-organizing which adds the
weighted differences between two population vectors to a third vector, therefore no separate
probability needs to be used (Adeyemo and Enitan, 2011).
Differential evolution algorithm is a stochastic optimization method minimizing an objective
function that can model the problem's objectives while incorporating constraints. It can be
used for optimizing functions with real variables and many local optima (Pierreval et al.,
2003). There are different improved versions of original differential evolution that have been
reported in the literature, such as hybrid differential evolution (HDE) (Tsai and Wang, 2005),
Pareto differential evolution approach (PDEA) (Madavan, 2002), multi-objective differential
evolution algorithm (MDEA) (Adeyemo and Otieno, 2009b), multi-objective differential
evolution algorithm (MODEA) (Ali et al., 2012) and a combined Pareto multi–objective
differential evolution (CPMDE) (Olofintoye et al., 2014).
However, only few studies have been reported on the applications of evolutionary algorithms
in the optimization of anaerobic reactor for CH4 production. Recently, artificial neural
network coupled with genetic algorithm (ANN-GA) has emerged as one of the most efficient
methods in empirical modeling and optimization, particularly for non-linear systems (Abu
Qdais et al., 2010). Once an ANN-based and mass balance-based process model with fairly
good generalization capability is constructed, its input space can be optimized appropriately
to secure the optimal values of process variables.
Modelling and optimization of biogas production on mixed substrates of banana stem, cow
dung, saw dust, paper and rice bran waste, using ANN coupled with GA was reported
(Gueguim Kana et al., 2012). In another study, simulation and optimization of the effect of
operational pH, temperature, total volatile solids (TVS) and total solids (TS) on biogas yield
using ANN and GA was studied by Abu Qdais et al. (2010). The integration of the ANN
model with GA as the optimization tool resulted in identification of the optimal operational
digester parameters that lead to increase in CH4 yield by 6.9%. The study demonstrated that
optimization of models using evolutionary algorithms such as GA will help in better
36
prediction of process output such as biogas production, especially CH4 yield (Abu Qdais et
al., 2010).
A mathematical model of a laboratory-scale plant for slaughterhouse effluents biodigestion
for biogas production was formulated with the objective of obtaining a liquid effluent of low
COD and to generate CH4 as a byproduct, stored and then used as an energy source (Martinez
et al., 2012). Parameters of this model were fitted into a two-step algorithm. The authors first
adjusted the parameters that are directly related to the measured variables using GA, while a
gradient descendent algorithm was used for fine adjustment of all the whole parameters to
optimize for maximum CH4 yield. The results reported showed that the model used was able
to predict four system variables and CH4 generation (Martinez et al., 2012).
Wei and Kusiak (2012) used a data-driven approach for optimization of a biogas production
process in a wastewater treatment plant. A multi-layer perceptron neural network was used
for the construction of an optimization model. High computational complexity of the model
led to the use of a particle swarm optimization (PSO) algorithm to maximize biogas
production, by finding the optimal settings of controllable variables. The model solution has
resulted in an increase in biogas production under an optimized operational condition using
the PSO algorithm. However, two major challenges in this field of parameter estimation of
non-linear dynamic biological systems were numerical integration of differential equations
and finding global parameter values (Tsai and Wang, 2005).
The particle swarm optimization technique was used to identify the parameters for an offline
estimation of yield and kinetic coefficients in a non-linear dynamical model for anaerobic
wastewater treatment bioprocesses (Sendrescu, 2013). The identification scheme was
formulated based on the optimization problem. The error between the simulated response of a
parameterized model and the actual physical measured response of the system was optimized.
The multimodal function and classical iterative methods used in their study as reported, failed
due to its inability to find global optimum solutions. However, the parameters estimation for
the system was achieved by the PSO algorithm for the minimization of error function. The
37
Table 2.2: Anaerobic model and optimization tools for different types of wastewater
System modelled and
wastewater types
Model type Evolutionary
algorithm type
Input Output References
UASB reactor treating
poultry wastewater
Empirical - HRT, OLR Daily biogas production rate, effluent
COD concentration.
Yetilmezsoy and
Sakar (2008)
UASB reactor treating
potato wastewater
- - pH, influent and effluent
COD, temperature, VFA,
alkalinity
Biogas production rate. Barampouti et al.
(2005)
Anaerobic hybrid reactor
for treating alkali rice
straw
Stover-Kincannon - OLR, HRT Biogas production and effluent COD
concentration.
Narra et al. (2014)
Distillery wastewater Stover-Kincannon - OLR, HRT Effluent concentration. Acharya et al. (2011)
Olive mill solid waste Chen-Hashimoto - Substrate Concentrate, HRT Volumetric methane production rate. Borja et al. (2003)
Poultry wastewater
treatment in UASB reactor
Chen-Hashimoto
and Stover-
Kincannon
LOQO/AMPL
algorithm
Influent concentration,
temperature, OLR, HRT,
reactor volume and flow
rate
Methane production rate, effluent
substrate concentration and net operating
cost
Yetilmezsoy (2012)
Dry–thermophilic
anaerobic digestion of
organic fraction from
municipal solid waste
Model based on
Romero García
(1991)
- Influent and effluent
concentration, HRT
Methane production. Fdez-Güelfo et al.
(2012)
Biogas production digester - ANN-GA Temperature, TS, TVS, pH Methane production and biogas yield. Abu Qdais et al.
(2010)
Real cotton textile
wastewater treatment in
UASB reactors
- ANN HRT, influent COD
concentration, pH, VFA
concentration, operating
temperature, dilution ratio,
alkalinity, TSS and OLR
COD removal efficiency. Yetilmezsoy and
Sapci-Zengin (2009)
38
strategy of PSO algorithm could still converge to accurate results, even in the presence of
measurement noise. The authors reported that PSO algorithms can be used in the optimization
of parameters during model identification (Sendrescu, 2013). Further studies on anaerobic
model and optimization tools for different types of wastewater are summarized in Table 2.2.
Models applied to describe AD have been shown to be an effective tool and could be used in
the near future for tracking and predicting the development of biogas production especially
CH4 yield. This could help in accelerating the speed of digester start-up, as well as biogas
(CH4) production from biomass and wastes for energy generation. Likewise, data-base
system could be developed for the simulated results in the field of renewable energy for
gathering and sharing information on the suitable technology, provide an appropriate
operational conditions for AD of wastes and thus, improve the amount of biogas formation.
2.8 RESEARCH OUTPUT
Journal Articles
1) Adeyemo, J. and Enitan, A. 2011. Optimization of fermentation processes using
evolutionary algorithms. Scientific Research and Essays, 6(7):1464-1472.
39
CHAPTER THREE: PERFORMANCE EVALUATION OF AN UPFLOW
ANAEROBIC SLUDGE BLANKET REACTOR TREATING BREWERY
WASTEWATER
3.1 INTRODUCTION
The brewing process involves series of batch operations on raw materials to the final product.
The production process includes the blending and fermentation of maize, malt and sorghum
grits using yeast, which requires large volumes of water as the primary raw material.
Traditionally, the amount of water needed to brew beer has been found to be several times the
volume actually brewed (Simate et al., 2011). For instance, an average water consumption of
6.0 hectolitres is required to produce one hectolitre of clear beer (South African Breweries,
2001). Large volumes of water are being used by the industry for production of beer for two
distinct purposes; as the main ingredient of the beer itself and as part of the brewing process
for steam raising, cooling, washing of floors, cleaning of the brew house, packaging and
cleaning after the completion of each batch operation. The amount of wastewater that is being
discharged from the industry after the production of beer also contributes to this large volume
of water (Simate et al., 2011). This wastewater is very high in organic content and is highly
polluting to the environment if discharged without prior treatment (Raposo et al., 2010;
Inyang et al., 2012; Mata et al., 2012). Furthermore, most industries discharge their effluents
into municipal treatment plants or to the environment without adequate characterization,
quantification and pre-treatment due to economic and technological constraints (Ikhu-
Omoregbe et al., 2005). This may have an adverse effects on the municipal treatment plants
by overloading these systems thus, reducing the efficiency of the treatment plants.
Among the brewery industries, South African Breweries (SAB Ltd) has been reported to be
the second largest beer producer in the world and uses about 10.5 million cubic metres of
water per annum at one of its breweries and approximately 70% of this is discharged as
wastewater (Jones et al., 2011; Kirin Holdings, 2012). With the competing demand on water
resources and water reuse, discharge of industrial effluents into the aquatic environment has
become an important issue (Islam et al., 2006; Danazumi and Hassan, 2010; Kanu and Achi,
2011; Simate et al., 2011; Kovoor et al., 2012). Much attention has been placed on the impact
of industrial wastewater on water bodies worldwide due to the accumulation of organic and
40
inorganic suspended matter, nitrite, nitrate and soluble phosphorus (Phiri et al., 2005; Islam
et al., 2006; Baig et al., 2010; Ipeaiyeda and Onianwa, 2012).
A considerable number of environmental pollution problems have emerged recently, which
has led to monitoring and controlling of the quality and quantity of liquid effluents being
discharged into natural water bodies or municipal treatment plants especially by the industry
(Kanu and Achi, 2011; Kovoor et al., 2012). The effects of contamination on water bodies
include change in pH, electrical conductivity, temperature and eutrophication of rivers and
dams due to high concentrations of inorganic and organic matter from the industrial activities.
Some industries have been fined by the national water authorities and municipal authorities
for discharging poor quality effluents that do not meet the discharge standards into the natural
water bodies, as well as the municipality water treatment plants (Ikhu-Omoregbe et al., 2005;
Parawira et al., 2005; Worldwide Brewing Alliance, 2011). In order to meet regulatory
standards, many industries including brewery industries pre-treat their effluent using different
AD technologies before its release into municipal treatment plants (Parawira et al., 2005;
Melamane, 2007).
High concentrations of pollutants load in the brewery wastewater are greatly reduced by the
use of high-rate anaerobic digestion (AD) technology. It has helped the industry to comply
with stricter pollution control regulations, satisfy the search for greater efficiency and
improves effluent quality (Parawira et al., 2005; Li et al., 2011). AD process produces less
sludge than aerobic treatments, hence reducing effluent disposal costs. The UASB system has
successfully been used to treat different types of wastewater (Nery et al., 2001; Manhokwe et
al., 2009).
In the last few decades, much attention has been paid to the use of AD processes for the
treatment of brewery wastewater due to the nature and strengths of the brewery wastewater
(Parawira et al., 2005; Nizami and Murphy, 2010). Benefits of using UASB reactors include
the production of sufficient amounts of biogas as a natural source of energy that can be used
as electricity to power the entire brewery wastewater treatment process (Bocher et al., 2008).
41
The aim of this study was to monitor, characterize and quantify the brewery wastewater
pollution load from one of the brewery industry in KwaZulu-Natal, South Africa. Thereafter,
the efficiency of a full-scale UASB reactor for the treatment of brewery wastewater and the
composition of biogas produced during AD was determined. This UASB reactor is being
used for on-site treatment of brewery wastewater to reduce the organic load before discharge.
This study will help in generating data for both the industry and the local authority, as well as
assess the level of compliance by the industry to the local legislative guidelines for effluent
disposal. The data obtained from the full-scale UASB reactor will also be used in the course
of this study to determine model coefficients to predict CH4 production and effluent quality.
3.2 MATERIALS AND METHODS
3.2.1 Description of Full-Scale UASB Reactor
The full-scale UASB reactor was constructed from concrete with a series of settlers and
baffle plates arranged at the bottom for even distribution with a pre-conditioning tank (Figure
3.1) and 20% effluent recycle. The operating capacity of this UASB reactor is 1480 m3
excluding the pre-conditioning tank (Ross and Louw, 1987); however the total capacity
increases up to 1700 m3
including the pre-conditioning tank (Isherwood, 1991). The pre-
conditioning tank is used to retain effluent for hours for solid settlement. The pre-
conditioning tank was used to homogenize the incoming effluent and balance the variations
in pH, organic loads and flow resulting from the brewing process operation to desired levels
of anaerobic treatment. The volume of the reactor was based on the average volumetric
loading rate of about 10 kg COD /m3 per day. Nitrogen from nutrient supplements are added
into the conditioning tank in the form of urea and FeCl3 to provide the biomass with
necessary nutrients for nitrogen and iron as well as to help in flocculation of the biomass in
the reactor. The adjustment of acidic influent to neutral pH is currently being done by the
addition of soda ash (Enitan et al., 2014a).
The operational temperature and pH of the reactor was maintained between 33 ± 2°C and
6.5-7.2 respectively. Retention time varies with influent flow rate between 8-12 h for bacteria
42
to make use of the influent substrate. The biogas produced in the reactor is separated from the
effluent and the biomass in three-phase separators at the top of the reactor. The gas passes
through a defoam tank to remove any solids present, and was then flared. The biomass
separated from the gas and effluent is retained in the reactor and settled back into the sludge
bed. Off gas produced at the surface of the weirs in the UASB reactor is currently collected
and treated through a biofilter prior to being vented to the atmosphere (Enitan et al., 2014a).
The treated UASB effluent is disposed to the municipal wastewater treatment plant for
further treatment.
3.2.2 Wastewater and Biogas Sampling Procedure
Raw brewery wastewater, influent (pre-conditioned wastewater) and effluent (post-reactor
treatment) wastewater samples were each collected in one-litre sterile glass bottles (Figure
3.2) and transported to the laboratory at 4°C for analysis. Physico-chemical analyses were
carried out within 48 hours of sample collection over a period of one year with necessary
preservation techniques adapted from Standard Methods (APHA–AWWA–WPCF, 1998).
Biogas was collected into a gas holder (Tedlar bag, Sigma-Aldrich) for analysis. At first,
sample was taken bi-weekly but changed to monthly basis from the third month
3.2.3 Wastewater Characterization
Wastewater samples were analyzed for total dissolved solids (TDS), total suspended solids
(TSS), volatile suspended solids (VSS), total solids (TS), volatile solids (VS), temperature,
pH, oxidation-reduction potential (ORP), alkalinity, total chemical oxygen demand (TCOD),
soluble chemical oxygen demand (SCOD), biological oxygen demand (BOD5), conductivity
(mS/cm), crude protein, sulphates, orthophosphate, total oxidised nitrogen (TON), nitrite
(NO2), nitrates (NO3) and NH3 according to Standard Methods for Examination of Water and
Wastewater (APHA–AWWA–WPCF, 1998). The TS and TSS were determined
gravimetrically by drying well homogenized samples respectively at 103°C for 24 h. The VS
and VSS fractions were determined gravimetrically by incineration in a muffle furnace at 550
°C for 1 h (APHA–AWWA–WPCF, 1998). Alkalinity was measured by potentiometric
titration using 0.02N H2SO4 to an end-point pH value of 4.5. The aim of measuring alkalinity
43
was to evaluate the buffering capacity of the UASB reactor treating brewery wastewater and
the effect on the granular sludge (APHA–AWWA–WPCF, 1998). Tests were carried out in
duplicate, means and standard deviations are presented where appropriate.
3.2.3.1 Conventional and instrumental methods used for analysis
The TDS, conductivity (mS/cm) and oxidation-reduction potential (O/R potential) were
measured using calibrated electrode (YSI 556MPS, Yellow Springs, USA). The pH and
temperature were measured using a pH meter (Beckman pH 211 Microprocessor, USA). The
BOD5 measurement was done using the respirometric method for five days (OxiTop TS
606/2-i system). The COD concentration in the wastewater was determined by close
refluxing according to the standard method, 5220D (APHA–AWWA–WPCF, 1998). The
protein concentration was analyzed using a UV/VIS Spectrophotometer (Merck,
Spectroquant Pharo 300, Germany) according to the protocol of Lowry et al. (1951).
Sulphates, orthophosphate, NH3, TON, NO2 and NO3 were measured using Thermo Gallery
photometric analyser (Thermo Scientific, UK) (APHA–AWWA–WPCF, 1998). The
composition of biogas produced was analyzed using a gas chromatograph (Shimadzu GC-
2014, Japan) equipped with a thermal conductivity detector (TCD). The column used was
Porapak Q 1.8 m × 2.10 mm with the column oven, injector and detector temperatures set at
40°C, 100°C and 100°C, respectively. Helium gas was used as the carrier at 20 ml/min.
44
Figure 3.1: Layout of full–scale UASB reactor treating brewery wastewater (Hoffmann,
1985; Ross, 1989).
45
Figure 3.2: Schematic diagram of the sampling points from which samples were collected for
this study to monitor the full-scale UASB reactor treating brewery wastewater.
3.2.4 Analytical Quality Assurance and Statistical Analysis
Both reagent and sample blanks were used for all the methods that required the use of the
spectrophotometer and Aquakem Gallery discrete auto analyser. Standard solutions were
prepared for the analysis of COD and protein content. Instruments were first calibrated before
using standard solutions. Sample bottles were thoroughly cleaned, 1:1 HCl, triple rinsed with
distilled water and a final triple rinse was done with the sample as suggested by Fatoki and
Mathabatha (2001). The data obtained was used to calculate mean, ranges, and standard
deviations. Graphs and statistical analysis were performed using GraphPad Prism v 5.0,
software package for Pearson‘s correlation coefficient and analysis of variance (ANOVA) of
the parameters measured.
Sampling point
Digester in
Biogas
UASB reactor
Sampling point
Digester out
To municipal works
Post-aeration tank
Sludge blanket
Flare
Gas buffer
Boiler house
for brewery
HCl NaOH
Static
Screen
Pre-
conditioning
tank Sampling point
for raw brewery wastewater
46
3.2.5 Estimation of Pollutant Removal Efficiency
The organic load, nutrient and suspended solid removal efficiency of the UASB reactor were
calculated using Equation 3.1 (Clara et al., 2005).
( )
(3.1)
Where, Cinfluent = initial parameter concentration and Ceffluent = final parameter concentration.
3.3 RESULTS AND DISCUSSION
3.3.1 Brewery Wastewater Composition
The results of the physico-chemical analyses and the summary of the statistical analysis of
the raw brewery wastewater composition investigated for over a period of one year are shown
in Table 3.1. The results showed that the effluent produced from the brewery industry did not
meet the discharge limit for wastewater disposal to water bodies according to the European
Union (EU) discharge limits (Driessen and Vereijken, 2003). Although, the local effluent
discharge standards do vary from one location, region and country to another, as shown in
Table 3.1 (Department of Public Works Republic of South Africa, 2012). Furthermore, the
discharge limits are less stringent when the effluent is to be discharged to a municipal
wastewater treatment plant (Adeniyi, 2002).
The results of the analysis indicated that the qualities of the raw brewery wastewater from the
industry prior to treatment in terms of total and soluble COD as well as the BOD5 are higher
than the discharge standards (Table 3.1). The trends and variability of the values as well as
large standard deviations from the means shows that the pollution level of the wastewater was
high. The average and standard deviation of the total and soluble COD values of wastewater
prior to treatment were 5340.97 ± 2265.11 mg/L and 3902.28 ± 1644.25 mg/L respectively.
The trends of total and soluble COD during the course of the brewery wastewater
composition monitoring showed fluctuation in the strength and composition of the brewery
wastewater with the range being between 1096.41 to 8926.08 mg/L for TCOD and 1178.64 to
5847.74 mg/L for SCOD (Enitan et al., 2014b). The variations in the COD concentration for
each week could be as a result of variation in the activities and housekeeping practices of the
47
brewery plant. The observed values are within the range reported for some brewery
wastewater plants as shown in Table 3.2 (Ikhu-Omoregbe et al., 2005; Parawira et al., 2005;
Rao et al., 2007; Inyang et al., 2012).
The variation in BOD5 and SCOD contents of the brewery wastewater. The BOD5 values
range between 1609.34 – 3980.61 mg/L with a mean value of 3215.27 mg/L and standard
deviation of ± 870.90 (Table 3.1). Low COD: BOD5 ratios of 1.932 ± 0.543 obtained in this
study were in accordance with past reports, which suggested that the wastewater content is
biodegradable (Kilani 1993; Dupont, Theodore and Ganesan 2000). Effluent from the
brewery plant is regarded as a biodegradable industrial wastewater and the COD
concentration of brewery effluent that is more than 800 mg/l has been reported to be more
suitable for treatment using anaerobic digestion (Dupont, Theodore and Ganesan 2000; Ikhu-
Omoregbe, Kuipa and Hove 2005). Further work on the characterization of brewery
wastewater during the monitoring period could be found in the literature (Enitan et al.,
2014b).
* All values are in mg/L except otherwise stated.
*An average of 14 samples ± std deviation.
Table 3.1: Summary of raw brewery wastewater composition from the industry prior to anaerobic
treatment and indicative discharge limits in South Africa (SA) and the EU (Driessen and Vereijken, 2003)
Parameters Range Average value* SA Discharge
limits
EU Discharge
Limits
Temperature (˚C) 24-30.5 27.90 ± 2.23 ˂ 44 -
pH 4.6-7.3 6.0 ± 1.44 Between 5.0 and 9.5 -
Total COD 1096.41- 8926.08 5340.97 ± 2265.11 75 125
Soluble COD 1178.64 - 5847.74 3902.28 ± 1644.25 - -
BOD5 1609.34 – 3980.61 3215.27 ± 870.90 Determined by the treatment capacity of the
receiving sewerage treatment plant 25
Crude protein 61.67-754.42 273.47 ± 233.63 - -
Orthophosphates 7.51 -74.10 23.71 ± 21.88 10 1-2
TON 0 - 5.36 1.81 ± 1.66 - -
NH3-N 0.48 - 13.05 8.62 ± 10.40 3 -
Nitrate 1.14 -11.55 4.30 ± 3.41 15 -
Nitrite 0-0.24 0.37 ± 0.18 15 -
ORP (mv) -27.10 to -84.91 -47.80 - -
Conductivity (mS/cm) 1.04-1.62 1.52 70-150 -
TS 1289.26 – 12248.13 5698.11±2749.06 - -
VS 1832.82 – 4634.31 3257.33± 1074.34 - -
TSS 530.67 – 3728.02 1826.74± 972.46 25 35
VSS 804.11 -1278.43 1090.86 ± 182.74 - -
Alkalinity(mgCaCO3/ L) 500- 10000 2450.33± 3034.19 - -
48
* All values are in mg/L except otherwise stated
3.3.2 Efficiency of UASB Reactor Treating Brewery Wastewater
3.3.2.1 Effect of pH and temperature on UASB reactor performance
Raw wastewater from the brewery industry was acidic and adjustment to neutral pH was done
by the addition of soda ash inside the conditioning tank prior to treatment. This was done
because the anaerobic reactor is very sensitive to changes in pH and if wastewater is not
buffered, it could lead to accumulation of VFA concentrations in the reactor and thus, affect
the activity of microorganisms (Rosenwinkel et al., 2005). The concentration of substrate in
the pre-conditioned brewery wastewater and the pollution effect on the treatment plants are
presented in Figure 3.3. This figure shows the observed pH values of the pre-conditioned
wastewater (digester inlet) and effluent from the reactor (digester outlet) with the
corresponding substrate COD concentrations. The reactor‘s pH was stable throughout the
monitoring period between 6.6 and 7.3 at an average operating temperature of 29˚C under
various organic loading rates.
Table 3.2: Brewery wastewater characterization and the efficiency of the UASB reactor as compared to the
literature
Parameter Units This study Parawira et
al. (2005)
Ahn et al.
(2010)
Rao et al.
(2007)
Diaz et
al.
(2006)
Rüffer et al.
(1991)
Inyang et
al. (2012)
pH - 4.6-7.3 3.30-6.30 6.3-6.9 3-12 7.2 - 11.97
Temperature ºC 24-30.5 25-35 - 18-40 - - -
NH4-N mg/L 0.48 - 13.05 - 2.2-7.0 - 15 - -
TN mg/L 0 - 5.36 0.0196-
0.0336
17-36 - 15 30-100 0.39
TP mg/L - 16-124 8.4-17 - - 10-30 0.462
COD mg/L 1096.41- 8926.08 8240 ≥ 20000 910-1900 2000-6000 4000 1120-1500 471
TSS mg/L 530.67 – 3728.02 2020-5940 140-320 2901-3000 1300 10-60ml/l 81
VSS mg/L 804.11 -1278.43 - 90-180 - - - -
TS mg/L 1289.26 –
12248.13
5100-8750 1300-2000 5100-8750 - - -
CODremoval % 78.97 57 80 - 80 - -
Total COD
quantity in
reactor
g 13210.48 10,000 - - - - -
Total COD
removal
g 10436.28 5700 - - - - -
49
Several studies have reported reactor failure or under performance of their anaerobic
treatment system due to low pH values and changes in reactor temperature (Visser et al.,
1993; Poh and Chong, 2009; Tabatabaei et al., 2011). In a study conducted by Tai et al.
(2006), a similar trend in the pH of effluent from the UASB reactor with pH values between
6.9 and 7.5 was reported. This condition is considered optimal for the growth of methanogens
(Gerardi, 2003). Steinhaus et al. (2007) studied the optimum growth conditions of
Methanosaeta concilii using a portable anaerobic microtank. They reported an optimum pH
level of 7.6 for the growth of methanogens and any deviation from this optimum pH could
lead to the inhibition of methanogens in the anaerobic reactor as well as CH4 production.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1000
2000
3000
4000
5
6
7
8
9
10
COD in pH-inlet pH-outlet
Time (weeks)
CO
D c
once
ntra
tion
(mg/
L)
pH
Figure 3.3: The effect of inlet COD variations on the pH of the full-scale UASB reactor
treating brewery wastewater.
However, in this study, it was observed that operational temperature affected the pH of the
reactor, which in turn determined the amount of biogas produced and CH4 content
respectively. Figure 3.4(a) presents the operational temperatures of the reactor against the
final pH values of the AD of brewery wastewater using the full-scale UASB reactor. A
simple linear regression was performed on the data to determine if there was a significant
relationship between pH and temperature. A poor positive relationship between the final pH
50
of the treatment unit and the operational temperature was shown by a low Pearson‘s
correlation coefficient of R = 0.177 (Figure 3.4b). The statistical result indicated that there
was a weak positive relationship between these two parameters.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1624
26
28
30
32
34
5.2
5.6
6.0
6.4
6.8
7.2
7.6
8.0
8.4
8.8T (C) Final pH(a)
Duration (weeks)
Tem
pera
ture
(C
)
Fin
al p
H
24 26 28 30 32 346.0
6.5
7.0
7.5
8.0
y=0.023x + 6.25 (R = 0.177)
(b)
T (oC)
Fina
l pH
Figure 3.4: (a) Change and (b) the relationship between reactor temperature and final pH
value of UASB reactor treating brewery wastewater
51
3.3.2.2 COD removal efficiency and solids concentration
In this study, the characterized raw brewery wastewater required additional nutrient nitrogen
for the anaerobic microorganisms due to a low COD: N ratio. Urea was added as additional
nitrogen. The UASB reactor was fed with pre-conditioned wastewater with an average COD
concentration of approximately 2005.73 ± 1139.85 mg/L at 28˚C. During the monitoring
period (Figure 3.5), the effluent from the UASB reactor had a considerably low level of COD
concentration remaining after treatment (457.25 ± 272.41 mg/L). COD removal efficiency
was 78.97% on average (Table 3.3).
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 150
1000
2000
3000
4000
5000
0
10
20
30
40
50
60
70
80
90
100
COD inletCOD outlet % COD removal
Duration (weeks)
CO
D c
onc
entr
atio
n (m
g/L
)
CO
D r
emo
val
(%)
Figure 3.5: Performance of the full-scale UASB reactor treating brewery wastewater in terms
of COD removal efficiency.
Ochieng et al. (2003) and Parawira et al. (2005) reported a high COD removal efficiency for
brewery wastewater enriched with nitrates and phosphates, compared to the wastewater
without enrichment. COD removal efficiencies ranging from 80 to 90% have been achieved
using different industrial effluents in UASB reactors (Britz et al., 2002; Manhokwe et al.,
2009; Atashi et al., 2010). In a similar study carried out by Kilani (1993), the effect of dairy
and brewery effluents on the treatment efficiency of a domestic sewage system was
investigated. An average COD removal of 60% was reported using a laboratory scale reactor.
Atashi et al. ( 2010) reported about 90% COD removal efficiencies from a pilot-scale UASB
52
reactor treating sugar mill wastewater. Table 3.2 showed earlier presented few examples of
brewery wastewater characterization studies from the literature and the efficiency of the
anaerobic treatment units in organic matter removal.
Table 3.3: Composition of influent (brewery wastewater after pre-conditioning) and UASB effluent
Parameters (Average values) Digester inlet Digester outlet % Decrease %
Increase
Temperature (˚C) 29.21 29.46 - -
pH 6.87 6.93 - -
COD 2005.73 457.25 78.97 -
TSS 2449.46 3268.97 - 33.46
TS 4520.10 3295.67 27.09 -
TDS 1792.80 2043.20 - 13.97
Protein content 134.40 71.39 46.88 -
Orthophosphates 21.25 25.34 - 19.21
TON 0.52 0.48 7.65 -
NH3 21.64 53.85 - 148.85
NO2 2.30 1.99 13.53 -
NO3 0.07 0.34 - -
Sulphate 178.25 826.28 - -
ORP (mv) -144.78 -73.15 - 42.89
Conductivity(mS/cm) 2.18 2.59 - 18.49
* All parameters are in mg/L except otherwise stated.
Figure 3.6 shows the values of TSS removal in the UASB reactor with an inlet and outlet TSS
concentration of the brewery wastewater. An increase in the effluent total suspended solids
was observed with an average increase of 33.46%. This shows that the discharged effluent
was higher in TSS concentration than the influent. Furthermore, there was a marked decrease
in COD removal efficiency and total solids at week 9, with low COD removal efficiency of
49.90% recorded (Figure 3.5). This might have been as a result of an increase in total
suspended solids (TSS) of the influent composition from 3063.41 mg/L to 11176.38 mg/L of
effluent from the reactor (Figure 3.6a). A second order quadratic polynomial regression
between %TSS and %COD removal showed a strong non-linear relationship with an R2
of
0.910 (Figure 3.6b). This could be attributed to the high concentration of the protein in the
influent before treatment. The protein can easily be converted to biomass which in turn
increases the reactor TSS, thus leading to sludge wash out from the reactor as shown in the
effluent TSS value. Structural problems in the 3 phase separator and effluent weirs could be
another contributing factor to biomass wash out.
53
0 1 2 3 4 5 6 7 8 9 10 110
2000
4000
6000
8000
10000
12000
-300
-200
-100
0
100TSS inlet TSS outlet % TSS(a)
Duration (weeks)
TS
S c
onc
entr
atio
n (m
g/L
)
% T
SS
rem
ova
l
20 40 60 80 100
-300
-200
-100
0
100
(b)
y = -0.4492x2 + 22.86x - 2704 (R2 = 0.910)
% COD removal
% T
SS
rem
ova
l
Figure 3.6: (a) Performance of the UASB reactor treating brewery wastewater in terms of
total suspended solids removal and (b) the second order quadratic polynomial regression
between %TSS and %COD removal efficiency of the UASB reactor.
54
Uyanik et al. (2002) and Zhou et al. (2006) mentioned the importance of extracellular
polymeric substances (EPS) in granule formation (Miksch and Beata, 2012). Taking into
account the results of the percentage degradation of COD, TSS, biogas and high biomass
formation in the UASB reactor studied, the results suggested that some of the biodegradable
COD is converted to biomass with biomass profile of 800-1000 ml rather than biogas
formation. This confirmed the problem often encountered in the treatment of brewery
wastewater (Zvauya et. al., 1994). Hence, it is very important to improve the performance of
this UASB reactor in terms of COD removal and TSS concentration in the final effluent.
3.3.2.3 Nitrogen and phosphate concentrations in the wastewater
There was variation in inlet and outlet concentrations of nitrite for the treatment of brewery
wastewater using UASB reactor. The nitrite-nitrogen load was reduced to 1.99 mg/L from an
influent concentration of 2.30 mg/L nitrite-nitrogen (Table 3.3). Very little residual nitrate
(0.34 mg/L) was detected in the effluent at an influent concentration of 0.07 mg/L, which
shows that there was an increase in nitrate-nitrogen in the reactor during organic matter
degradation, which in turn favours both the denitrification and methanogenesis processes
(Atashi et al., 2010). However, the performance of the reactor may be disturbed by the
increase in nitrate-nitrogen load, thereby having an inhibitory effect towards the anaerobic
biodegradation of biomass which in turn reduces the CH4 production (Sternenfels, 2012).
Thus, low nitrate concentration in this study might have contributed to high CH4 content
produced in the reactor.
However, the digester effluents still have considerable amounts of NH3 content. Many studies
have shown that free NH3 and not ammonium is responsible for inhibiting the methanogenic
activity during AD (Sawayama et al., 2004; Calli et al., 2005; Garcia and Angenent, 2009).
Ipeaiyeda and Onianwa (2012) explained that the presence of NH3 concentrations in the
effluent has its origin from the proteins and chitins contained in the brewery waste, because
most nitrogen in the waste are in the form of NH3 following the degradation of proteins and
amino acid (Inanc et al., 2000). Ouboter et al. (1998), further mentioned that almost all of the
proteins in brewery effluent is mineralised through the activity of proteolytic and deaminative
bacteria, initially hydrolysing protein to peptides and amino acids and finally by deamination
55
to ammonium (NH4). This explains the major source of NH3 in the effluent after treatment in
the UASB reactor.
The NH3 content of the influent and effluent of the UASB reactor during the monitoring
period is shown in Figure 3.7. The concentration increased by 148.85% from an influent
concentration of 21.64 ± 10.70 to 53.85 ± 21.08 mg/L on average (Table 3.3). This showed
that there was production of NH3–N during the treatment of brewery wastewater in the UASB
reactor. The release of NH3–N in the bioreactor during treatment of waste was also observed
by Inanc et al. (2000) and Govahi et al. (2012). Furthermore, the NH3 concentration detected
in this study can be said to be within an acceptable level for the growth of the methanogenic
bacteria and biogas production (Tabatabaei et al., 2011). However, excess concentrations of
free NH3 can inhibit these microorganisms. Tabatabaei et al. (2011) reported a wide range of
total NH3 nitrogen concentrations that can inhibit the growth of methanogens.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 150
20
40
60
80
100
120
140
160
180NH3 inlet
NH3 outlet
Duration (weeks)
Am
mo
nia
conc
entr
atio
n (
mg/
L)
Figure 3.7: Variation in average inlet and outlet concentrations of ammonia nitrogen during
anaerobic treatment of brewery wastewater using the UASB reactor.
56
The average influent and effluent orthophosphate concentration during anaerobic treatment of
brewery wastewater by the UASB reactor is shown in Figure 3.8. The reactor was fed with an
average influent orthophosphate concentration of 21.25 ± 9.30 mg/L that was increased to
25.34 ± 11.21 mg/L. This shows that orthophosphate was produced during the degradation
process. Parawira et al. (2005) reported low removal efficiency of nitrogen and phosphorus
during the AD of brewery wastewater in a UASB reactor. This shows that phosphate was
released by the microorganisms in the reactor during the anaerobic wastewater treatment
process.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 150
10
20
30
40
50
60
70
80
90
0
10
20
30
40
50
60
70
80
90
100PO4 inlet (mg/L) PO4 outlet (mg/L)
Duration (weeks)
PO
4 in
let
PO
4 ou
tlet
Figure 3.8: Average orthophosphates concentration in the reactor during treatment of
brewery wastewater.
3.3.2.4 Correlation between methane production and operational variables
Methanogenesis was active in the reactor during the treatment of brewery wastewater, which
is shown by the efficiency of the UASB reactor in terms of biogas production with CH4
content of 60-69% of the total gas throughout the study (Figure 3.9). The relationship
between the percentage COD removal and biogas yield (CH4 and CO2) during anaerobic
degradation is shown in Figure 3.9. The results of analysis carried out using ANOVA showed
that biogas yield depended on the substrate present in the wastewater in terms of COD
57
removal efficiency (Appendix 1). There was a strong positive correlation between the
percentage COD removal and biogas yield (CH4 and CO2 production) with an R value of
0.975; which showed that significant portion of the organic matter presence in the brewery
wastewater was converted to biogas (Figure 3.9). The performance of the reactor as a
function of quantity of COD per reactor volume showed that high concentration of organic
matter in the reactor was used for biogas production as shown in Figure 3.9. Apart from the
composition of biogas produced, the quantity of COD removed in term of COD removed per
reactor volume shows the efficiency of the investigated full-scale UASB reactor during the
monitoring period.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 150
20
40
60
80
100
1000
2000
3000
4000
5000
6000
7000
8000
%CO2 %CH4 COD quantity in the reactor (g)
COD removal per reactor volume (g)% COD removal
Duration (weeks)
% o
f B
iogas
fo
rmat
ion a
nd C
OD
rem
ova
l
Quan
tity
of
reac
tor
CO
D a
nd i
ts r
em
ova
l (g
)
Figure 3.9: Efficiency of organic matter removal (COD quantity) as function of reactor
volume to produce biogas during anaerobic treatment of brewery wastewater.
As shown in Figure 3.10, CH4 gas production rate increased from 6.32 L/day to 19.47 L/day
as the OLR increased from 2.10 L/day to 9.30 g/L/day in UASB reactor; however an increase
in loading rates above 11.00 g COD/L/day resulted in a decrease in CH4 production to 15.43
L/day. Habeeb et al. (2011) and Govahi et al. (2012) earlier reported a negative correlation
between CH4 production rate and OLR when the latter was increased. Furthermore, increase
in OLR could cause VFA accumulation, which in turn could result in decrease in pH of the
58
reactor, thus inhibiting CH4 production (Habeeb et al., 2011). The result from the comparison
of the final pH of effluent from the UASB reactor treating brewery wastewater and the CH4
content of the biogas produced from this reactor shows that there was a moderate positive
correlation (R = 0.664) between these two parameters [Figure 3.11 (a and b)]. Thus, the pH of
the reactor had a significant effect on the CH4 production (P < 0.001). This is because CH4
producing Archaea or methanogens are known to be affected by pH (Poh and Chong, 2009;
Habeeb et al., 2011) and they could only survive a very narrow pH range as discussed in
section 2.5.6 (Gerardi, 2003; Tabatabaei et al., 2011).
0 2 4 6 8 100
5
10
15
20
25
Organic loading rate (OLR) (gCOD/L/day)
Met
hane
pro
duct
ion
rate
(L m
etha
ne/g
/d)
Figure 3.10: Effect of organic loading rate on methane production rate in the UASB reactor
treating brewery wastewater.
59
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1550
55
60
65
70
75
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0(a) % CH4 Final pH
Duration (weeks)
CH
4 co
nten
t (%
)
Fina
l pH
55 60 65 70 75
6.0
6.5
7.0
7.5
8.0
y = 0.052x + 1.24 (R = 0.664)
(b)
% Methane
Fin
al p
H
Figure 3.11: Graph showing (a) the effect of reactor‘s pH on the methane content and, (b) the
relationship and linear regression analysis showing a significant moderate positive correlation
between these two parameters during the treatment of brewery wastewater using the UASB
reactor.
60
3.4 CONCLUSIONS
Raw brewery wastewater characterization results showed that the process wastewater
from the brewery industry was high in COD, BOD5, TSS, NH3 and protein content and
did not meet the required effluent regulatory standards during the period of sampling.
From the results obtained in this study, the BOD5: COD ratio indicated that the raw
wastewater was high in organic matter which is biodegradable. Therefore, this effluent
could easily be degradable by the microorganisms in AD technology.
The full-scale UASB reactor was able to reduce the organic content in the brewery
wastewater to a reasonable level (average 457.25 ± 272.41 mg COD/L in effluent), which
could be discharged into the municipal wastewater plant for further treatment. However,
the results of the percentage removal efficiency of NH3 and phosphorus showed high
concentrations of these nutrients in the final effluent; therefore secondary treatment was
highly recommended.
The composition of biogas, especially the CH4 yield, showed that considerable amounts
of substrate was being converted to biomass due to an increase in the concentration of
total suspended solids and total dissolved solids in the final effluent. A very strong non-
linear relationship between the percentage solids and COD removal was observed. The
CH4 yield as a percentage of total biogas was between 60-69%. This showed that the
performance of the UASB reactor in terms of biogas (CH4) production could be improved
for energy generation, since CH4 production depends on the rate of organic matter
degradation.
As observed in this study, the UASB reactor needs optimization to improve the treatment
efficiency and post treatment of the final effluent is required for nutrient removal, in order
to meet the discharge standards. Also, the microbial population structure within the
anaerobic digester need to be investigated in order to determine the contribution to effluent
treatment and biogas production. Further optimization using mathematical models to
improve the efficiency of the reactor was required.
61
3.5 RESEARCH OUTPUTS
(a) Journal Articles
1) Enitan, A. M., Swalaha, F. M., Adeyemo, J. and Bux, F. 2014b. Characterization of
brewery effluent composition from a beer producing industry in KwaZulu-Natal, South
Africa. Fresenius Environmental Bulletin, 23(3): 693-701.
(b) Conference Papers
1) Enitan, A. M., Swalaha, F. M., Adeyemo, J. and Bux, F (2014). Evaluation of effluent
composition from a beer producing industry in South Africa. Presented at the International
Journal of Arts & Sciences’ (IJAS) American Canadian Conference at Ryerson
University‘s International Learning Center, Toronto, Canada, 19-22 May, 2014 (Oral
presentation).
2) Swalaha, F.M., Enitan A.M. and Bux, F. (2014). Efficiency of industrial scale anaerobic
reactor treating brewery wastewater. Presented at Water Institute of Southern Africa
(WISA) Conference, Mbombela, Mpumalanga, South Africa, 25-29 May, 2014 (Oral
presentation).
62
CHAPTER FOUR: KINETIC MODELLING AND CHARACTERIZATION OF
THE MICROBIAL COMMUNITY PRESENT IN AN UASB REACTOR
TREATING BREWERY EFFLUENT
4.1 INTRODUCTION
The anaerobic breakdown of complex organic compounds involves the action of several
groups of microorganisms which results in a variety of intermediates including biogas such as
H2, CH4 and CO2 (Appels et al., 2008; Mirzoyan et al., 2008; Amani et al., 2011). Microbial
species involved in the conversion of organic material in anaerobic digesters are grouped
based on their biochemical activities. This group includes hydrolytic, acidogenic, acetogenic
and methanogenic organisms (Hulshoff-Pol et al., 2004). These organisms grow in a
syntrophic manner when the digester is operated under optimum reaction conditions
(Chulhwan et al., 2005; Crocetti et al., 2006; Mumme et al., 2010). Studies have shown that
the microbial community in the UASB reactor responds to any sudden change in the
environmental conditions, thus leading to a shift in the type of microbial species found in the
reactor, their population size and activities (McHugh et al., 2003a; Diaz et al., 2006; Keyser
et al., 2006; Zhang et al., 2012). Therefore, an in-depth understanding of the microbial
consortia and the associated activities are essential for an effective reactor operation.
It is difficult to assess the diversity, colonization and topological distribution of these
microorganisms using conventional methods due to the structural complexity of the granular
sludge (Liu et al., 2002b). Recently, molecular techniques such as denaturing gradient gel
electrophoresis (DGGE), fluorescence in-situ hybridization (FISH), quantitative polymerase
chain reaction (QPCR) and pyrosequencing have been successfully adopted to study these
complex microbial populations (McHugh et al., 2003a; Diaz et al., 2006; Keyser et al., 2006;
Ziganshin et al., 2011; Zhang et al., 2012).
Furthermore, development and use of suitable mathematical models, which adequately
describe the overall process performance in the bioreactor have shown to be an important tool
for process control strategies resulting in better effluent quality and biogas production (Pontes
and Pinto, 2006). Mass balances, kinetic and stoichiometric models are some of the methods
that are being employed in describing the operating principles of different anaerobic digesters
63
(Acharya et al., 2008; Yetilmezsoy, 2012). Simple and more sophisticated models such as the
Monod, Chen and Hashimoto, Contois, Michaelis-Menten, Haldane, Grau second-order and
anaerobic digestion model 1 (ADM1) have also been developed to improve the reactor
performance (Batstone et al., 2002; Parsamehr, 2012).
Kinetic modelling is an acceptable method to describe and predict the performance of any
biological treatment unit (Yetilmezsoy and Sakar, 2008; Debik and Coskun, 2009). It can be
applied to the optimization and control of anaerobic wastewater treatment processes, to
determine the relationship between fundamental parameters needed for anaerobic reactions
(Acharya et al., 2011; Yetilmezsoy, 2012). Among several kinetic models developed for
organic substance removal in the UASB reactor, the Stover- Kincannon model has been well
documented (Acharya et al., 2011; Yetilmezsoy, 2012). The modified form of this model is
one of the most widely adopted methods for the determination of kinetic constants and has
been successfully applied for anaerobic treatment of poultry slaughterhouse waste (Debik,
and Coskun, 2009), municipality wastewater (Turkdogan-Aydinol and Yetilmezsoy, 2010),
distillery wastewater (Acharya et al., 2011) and poultry manure wastewater (Yetilmezsoy,
2012).
Thus, monitoring of environmental conditions and identification of the functional microbial
population, as well as analysing the kinetic process of UASB reactors is crucial for reactor
design, maintenance and its efficient operation to increase CH4 production as a source of
renewable energy and for better effluent quality. This chapter focused on determining and
quantifying the microbial composition in the granules collected from the full-scale UASB
reactor treating brewery wastewater in KwaZulu-Natal, South Africa using FISH, PCR and
QPCR techniques to detect and quantify the Bacteria and Archaea concentrations in the
reactor samples. The bio-kinetics of the degradable organic substrates present in the brewery
wastewater using Stover-Kincannon kinetic model to predict the effluent quality was further
considered. It is hoped that the characterization of eubacteria and methanogenic Archaea in
the granules used for this study will bridge the gap of knowledge on the microbial ecology of
the full-scale UASB reactor investigated.
64
4.2 MATERIALS AND METHODS
4.2.1 Sample Collection from the Full-Scale UASB Reactor
Well-suspended granular samples were obtained for microbial analysis from the UASB
reactor compartments as shown in the flow diagram (Figure 4.1). Prior to sample collection,
the sampling valves were opened for 5 minutes in order to flush out the sampling tubes and
valves. Thereafter, granular sludge samples were collected in sterile glass bottles and flushed
with nitrogen gas and sealed immediately to maintain anaerobic conditions during
transportation to the laboratory. Both granular sludge samples and wastewater samples
collected were transported to the laboratory at 4°C for analysis. Physico-chemical analyses
were done within 48 hours of collection with necessary preservation techniques adapted from
Standard Methods for Examination of Water and Wastewater (APHA–AWWA–WPCF,
1998). For microbial analyses, the aliquots were centrifuged at 9,600 x g at4˚C for five
minutes. Supernatants were discarded and the pellets were washed with phosphate buffered
saline (1 x PBS) and stored at -20˚C before analysis.
Volatile fatty acids (VFA) (acetic, propionic, isobutyric, butyric, valeric and isovaleric acids)
were quantified using HPLC (Model LC-20AT, Shimadzu, Japan) equipped with a UV
detector (SPD-20A) and analysed using a Metrosep organic acid column (250×7.8 mm) at a
flow rate of 0.6 ml/min and an injection volume of 20 μl at 210 nm. The mobile phase
consisted of a 0.5 mM H2SO4 solution. Biogas was collected in a gas holder (Tedlar bag,
Sigma-Aldrich) for analysis using gas chromatography (Shimadzu GC-2014, Japan) as
described in section 3.2.3.1.
Figure 4.1: Flow diagram showing the six sampling points from the UASB reactor
compartments where granular samples were obtained for microbial analysis.
Wastewater sample
Conditioning tank
Municipal treatment plant
Settling Tank
C6
C5
C4
C2
C3
C1
Compartment 1-6 UASB Reactor
65
4.2.2 Fluorescence In-Situ Hybridization (FISH)
Fluorescence in-situ hybridization was carried out according to the protocol described by
Amann et al. (1995) with minor modifications using the oligonucleotides probes given in
Table 4.1 (Enitan et al., 2014b). Sludge granules were fixed in 4% paraformaldehyde (Gram
negative) and in PBS-ethanol (Gram positive) (Amann et al., 1995). Fixed samples were then
sonicated at 2 W for 5 minutes using an Ultrasonic Liquid Processor (Misonix XL-2000
Series). Thereafter, granules were treated with 10 µl of lysozyme (10 mg/ml) at 37°C for 30
minutes; then with Proteinase K (10 mg/ml) at 50°C for 45 minutes. Samples were diluted
further by the addition of 500 µl of sterile water for even dispersion and quantification with
the group specific, Archaea and bacteria domain probes (Table 4.1). A volume of 5–10 µl of
the treated samples were fixed on poly-L-lysine coated slides and allowed to air-dry at room
temperature and dehydrated by a series of ethanol washes (50, 80 and 100%). The
oligonucleotide probes were labeled with rhodomine (FAM) and tetramethylrhodomine-5-
isothiocyanate (TAMRA) dye at the 5'-end respectively (Table 4.1). The hybridisation and
wash buffers were prepared according to the formamide stringency as listed in Table 4.1.
Samples were hybridised by the addition of 9 μl of hybridisation solution (10% SDS, 1 M
Tris/HCl (pH 8), 5 M NaCl and formamide concentrations; Table 4.1); together with 1 µl of
oligonucleotide probe, incubated in the hybridisation oven at 46°C, overnight. After
hybridisation, the slides were washed with pre-warmed wash buffer (1 M Tris/HCl, 10%
SDS, 0.5 M EDTA and 5 M NaCl; Table 4.1) for 1 h at 48°C; subsequently rinsed with
distilled water and then air dried. The slides were counter-stained with 4´-6-diamino-2-
phenylindole (DAPI) for 10 minutes at room temperature. Slides were rinsed in pre-warmed
distilled water and air-dried in the dark. The samples were then mounted with an anti-fading
solution (Vectashield, Vector Laboratories, Inc. Burlingame).
4.2.2.1 Microscopy and image analysis
The hybridized slides were viewed using a Zeiss Axio-Lab HB050/AC microscope (Carl
Zeiss, Jena, Germany) equipped with an HBO 50W Hg vapour lamp, with appropriate filter
sets, specific for TAMRA (43 HE, Zeiss) and FAM (Filter set 09, Zeiss) using a 100x Plan
Apochromat objective. Images were captured using a Zeiss AxioCam MRC camera (Carl
Zeiss, Gottingen, Germany) and analysed using the Zeiss Axio Vision Release 4.8 imaging
system.
66
a = Rhodomine, b = Tetramethylrhodomine-5-isothiocyanate.
4.2.3 Total Genomic DNA Extraction from Granular Sludge Samples
The full-scale UASB reactor has six different compartments (C1, C2, C3, C4, C5 and C6)
and samples were taken from each for microbial analysis. The direct isolation of total
genomic DNA from granular sludge samples was carried out using a phenol extraction
method (Sekiguchi et al., 1998; Klocke et al., 2007) with modifications. Two millilitres of
the sample were centrifuged at 9,600 x g for 5 minutes to release the microorganisms
entrapped within the granules and other undigested particles; after which the supernatant was
discarded. The pellets were washed twice with 1 x PBS and centrifuged again at 9,600 x g for
5 minutes to collect the pellets. Total genomic DNA was recovered by lysis of the cells by
adding 700 µl lysis buffer (0.5mol-1
EDTA, 0.1 mol-1
NaCl, 0.5 mol-1
Tris/HCl at pH 8.0)
with 0.2% ß-mercaptoethanol and 30-40 mg PVPP, then homogenised with 0.6 g sterile glass
beads at 600 x g for 5 minutes using bead beater machine. The granules were further treated
with 20 µl of Proteinase K (10 mg/L), vortexed to mix and incubated for 30 minutes at 37°C.
The suspension was incubated for 2 h at 65°C. Thereafter, the cells were freeze-thaw in series
(in dry ice: ethanol slurry for three minutes) and thawed at 65°C in a water bath for three
minutes. After cell lysis, debris, RNA and proteins were separated from the aqueous phase
containing the DNA by performing two-step phenol–chloroform–isoamyl alcohol extraction
(25:24:1) followed by 24:1 chloroform–isoamyl alcohol and centrifuged for 10 minutes at
Table 4.1: 16S rRNA oligonucleotide probes with the corresponding formamide stringency and NaCl
concentrations used in this study
Target group
Oligonucleotides
Probe name
Probe sequence (5′-3′)
Formamide
concentration (%)/ NaCl
(µl)
References
Archaeaa
ARC915 GTGCTCCCCCGCCAATTCCT 30 / 1020 Stahl and
Amann (1991)
Methanosarcinaa MS821 CGCCATGCCTGACACCTAGCGAGC 40 / 460 Raskin et al.
(1994)
Methanosaetaa MX825 TCGCACCGTGGCCGACACCTAGC 50 / 180 Raskin et al.
(1994)
Eubacteriab EUB338 GCTGCCTCCCGTAGGAGT 30 / 1020 Amann et al.
(1990)
EUB338 II GCAGCCACCCGTAGGTGT 30 / 1020 Daims et al.
(1999)
EUB338 III GCTGCCACCCGTAGGTGT 30 / 1020 Daims et al.
(1999)
67
13,800 x g to remove the phenol; this step was repeated until a clean interface was seen.
Precipitation of genomic DNA was done by the addition of 1 x volume of isopropanol and
stored at -20°C overnight for complete precipitation. The DNA was collected by
centrifugation at 13,800 x g for 20 minutes, washed twice with 90% ice-cold ethanol
followed by 70% ice-cold ethanol, air dried and dissolved in 100 µl TE buffer (0.5 mol-1
EDTA, 1 mol-1
Tris/HCl at pH 8.0). The concentration of the DNA was checked by
Nanodrop (ND-1000) Spectrophotometer. The purified DNA were stored at -20 °C and used
for further construction of the 16S rDNA clone library. DNA extraction was carried out in
duplicate for the UASB granules collected.
4.2.4 Amplifications using Polymerase Chain Reaction (PCR)
Polymerase chain reaction amplification conditions were optimized for methyl coenzyme-M
reductase gene (mcrA), domain Archaea, (ARC) and bacterial (BAC) genes using the
corresponding primer sets as listed in Table 4.2 (Giovannoni, 1991; Chan et al., 2001; Luton
et al., 2002). The PCR mixture contained 25 µl reaction volume of 0.3 µl of Taq DNA
polymerase (5 U/ml), 2.5 µl of PCR reaction buffer, 1 µl of each of the primers (10 mM), 0.5
µl of dNTPs (10 mM), 2 µl of the extracted DNA (10 µl) and PCR-grade water. The
modified PCR amplification conditions of Luton et al. (2002) was used as follows: initial
denaturation was performed at 94°C for 5 minutes; followed by 40 cycles of denaturation at
92°C for 1 minute; primer annealing at 52°C for 1 minute for mcrA and 53°C for 1 minute
(Archaea and bacteria), elongation at 72°C for 1 minute and a final extension was performed
at 72°C for 5 minutes. The PCR amplification was carried out in an automatic thermal cycler
Veriti (Applied Biosystems).
4.2.4.1 Agarose gel electrophoretic detection of PCR products
The PCR amplified products were resolved on 0.8-1.0% (w/v) agarose-Tris-borate EDTA gel
(ABgene, UK) (10 mM Tris-HCl, 10 mM boric acid, 2.5 mM EDTA, pH 8.0), visualized and
photographed under the BioDoc-It transilluminator system. An appropriately sized marker (1
kb DNA smart ladder) was included on each gel as a standard. Purification of the PCR
products was carried out for subsequent cloning with a commercial kit following the
manufacturer‘s instructions.
68
4.2.4.2 Cloning
4.2.4.2.1 Preparation of competent cells, ligation, transformation and clone analysis using
colony PCR
Escherichia coli DH5-α was selected as the competent cells for ligation. A single colony of
E. coli DH5-α was subcultured from the stock plate onto prepared Luria Bertani (LB) agar
antibiotic agar plates supplemented with 50 mg/ml of Ampicillin and incubated overnight at
37˚C. A day before the transformation, 2 ml of LB antibiotic broth was seeded with single
overnight bacteria colony and incubated overnight at 37°C in a shaker. Selected PCR
amplicons were ligated into the pTZ57R/T vector using T4 DNA ligase of the insTAclone
PCR cloning kit (Invitrogen) according to the manufacturer‘s instructions (Thermo Scientific,
InsTAclone PCR Cloning Kit). Ligation and insertion were carried out in a 30 µl reaction
volume constituting of 6 µl of ligation buffer, 3 µl PCR product containing the DNA, and 3
µl digested plasmid DNA, 1 µl enzyme T4 DNA ligase and 17 µl nuclease-free water in an
Eppendorf tube. The mixture was briefly vortex and centrifuge for 3-5 s at 9,600 x g. This
was incubated overnight at 4°C. In a separate micro-centrifuge tube, 2.5 µl of the ligated
mixture was directly transformed into the prepared competent cells and incubated on ice for
five minutes. These were then, plated onto pre-warmed LB antibiotic agar plates containing
X-gal (20 mg/ml) and IPTG (100 mM) stock solutions and incubated overnight at 37°C using
standard procedures (Sambrook and Russell, 2001). White clones were randomly selected on
LB antibiotic agar plates containing X-gal and IPTG stock solutions and positive clones were
confirmed by colony PCR using appropriate primer-sets and resolved on agarose gel for
further confirmation of plasmids containing the targeted inserts.
69
Table 4.2: Primer sets used in this study for both conventional and quantitative real-time PCR
Target
group
Target
microorganism
Primer
name
Sequences(5'→3') Amplicon
length (bp)
References
16S rDNA Archaea ARC622f
ARC915r
TGAAATCYYRTAATCCC
GTGCTCCCCGCCAATTCCT
246-250 Chan et al. (2001)
McrA
Functional gene
for methanogenic
Archaea
MLf
MLr
GGTGGTGTMGGATTCACACARTAYGCWA
CAGCTTCATTGCRTAGTTWGGRTAGTT
464-491 Luton et al. (2002)
16S rDNA Bacterial 27f
1492r
AGAGTTTGATCMTGGCTCAG
TACGGYTACCTTGTTACGACTT
~1500 Giovannoni
(1991)
4.2.4.3 Sequencing and phylogenetic analysis
Positive clones from compartments 1, 3 and 6 of the six compartments were selected for
sequencing to assess organisms at the top, middle and the bottom of the reactor (Inqaba
Biotechnical Industries Laboratory, South Africa). The obtained bacteria, archeon and mcrA
gene sequences were manually edited and similarity searches for the DNA sequences were
carried out using the Basic Local Alignment Search Tool (BLAST) program to search in the
(http://www.ncbi.nlm.nih.gov/BLAST) National Centre for Biotechnology Information
(NCBI) sequence database. The nucleotides sequences obtained from the GenBank were
converted to amino acid sequences and then aligned in CLUSTAL X. The aligned amino acid
gene sequences were edited using BioEdit and exported to MEGA version 5.10. Evolutionary
analyses were conducted in MEGA version 5.10 software (Tamura et al., 2011). The
phylogenetic trees were constructed from the alignments and bootstrap analyses were
performed using 1000 replicates by the neighbour-joining method (Saitou and Nei, 1987).
4.2.4.3.1 Nucleotide sequence accession number for samples obtained from the full-scale
UASB reactor
The obtained nucleotide sequences for methyl coenzyme-M reductase gene (mcrA), domain
Archaea, (ARC) and bacterial (BAC) obtained from the full-scale UASB reactor treating
brewery wastewater were submitted to the National Centre for Biotechnology Information
70
website (NCBI) under the accession numbers KF715644–KF715648 for mcrA gene,
KM191135–KM191137 for Archaea and KM065733–KM065740 for bacterial clones.
4.2.5 Quantitative Real-time PCR
Quantification of gene copy numbers in the extracted DNA samples were performed using
real–time PCR machine (C-1000 Touch, CFX 96, Bio-Rad Laboratories Pty Ltd, USA) with
two primer sets targeting the Archaea and bacteria domain, Table 4.2 (Steinberg and Regan,
2009). For each reaction mixture, amplification was carried out in a final volume of 20 μl
containing 10 μl of the Sso fast Eva green Master Mix (Bio-Rad Laboratories Pty Ltd, USA)
1 μl of each primer (final concentration, 10 µM), 4 μl of template DNA and PCR-grade water
was added to a final volume of 20 µl. Two-step amplification of the target DNA were carried
out using the modified protocol described by Steinberg and Regan (2009) as follows: initial
denaturation for 3.5 minutes at 94°C followed by 40 cycles of 30 s at 95°C and annealing for
30 s at 55°C and final extension with image capturing at 72°C for 30 s. For melting curve
analysis, the temperature was increased at 0.5˚C every 10 s from 40 to 95˚C. Each QPCR
assay was conducted in duplicate. For all experiments, appropriate negative controls
containing no genomic DNA were subjected to the same procedure to exclude any possible
contamination or carry-over.
Standard curve was obtained by plotting quantification cycle (Cq) as a function of log of
copy number of target DNA. Standard curves were constructed from purified PCR amplicons
for Archaea and bacteria primers. Standard curve for bacteria was constructed from a series
of 10-fold dilution of target DNA using the primer sets of 27f and 1492r targeting the 16S
rDNA gene of bacterial at a concentration of 2.77 x 103 to 2.77 x 10
10 copies/ng DNA. A
second standard curve for Archaea was constructed from a series of 10-fold dilution of target
DNA using the primer sets of ARC622f and ARC915r targeting the 16S rDNA of the domain
Archaea at a concentration of 1.64 x 104 to 1.64 x 10
11 copies/ ng DNA.
For each QPCR assay, the value of the logarithmic starting quality for the different 16S
rDNA gene were plotted against the threshold cycle (Cq) numbers and the linear ranges of
71
the standard curves were selected based on the R2 of the slope greater than 0.990. For
quantification of 16S rDNA gene concentration that were present in the DNA obtained from
the different compartment, the Cq values for each sample were compared with the
corresponding standard curves. Equation 4.1 was used to calculate the target 16S rDNA gene
copy numbers in each sample (Yu et al., 2006; Tan et. al., 2013). An average molecular
weight of 660 Da with the 6.02 × 1023
Avogadro's numbers are assumed for a base pair in the
double-stranded DNA (He et al., 2003).
( ⁄ ) ( ⁄ ) ( ⁄ )
( ) ( )⁄ (4.1)
4.2.6 Kinetic Analysis Using Stover–Kincannon Model
According to the Stover–Kincannon model (Kincannon and Stover, 1982), the organic
substrate utilization rate in a UASB reactor process can be expressed as a function of organic
loading rate. The substrate consumption rate can be expressed as (Acharya et al., 2008;
Turkdogan-AydInol and Yetilmezsoy, 2010; Yetilmezsoy, 2012);
( ) (4.2)
The original Stover-Kincannon model is described in equation (4.2) as;
( )
(
)
(
) (4.3)
Where dS/dt is the substrate removal rate (g COD/L/day) in the UASB reactor, S is the
reactor substrate concentration (g/L), Umax is the maximum utilization rate constant (g/L/day),
Vr is the working volume of reactor (L), KB is the saturation constant (g/L/day), Q is the flow
rate (L/day), Si and Se are the influent and effluent substrate concentrations (g/L)
respectively.
72
Combining equation (4.2) and (4.3) gives the modified Stover- Kincannon model for a UASB
reactor at steady state.
(
)
( )
( )
(4.4)
Y = λX + λ0,
( ) ,
,
( ), λ0 =
Considering the mass balance of substrate present in wastewater that flows into the reactor
and out of the reactor plus the total amount of substrate degraded, at a specific flow rate,
control volume and time, then the mass balance can be written as;
(
) (4.5)
Substituting dS/dt from the equation (4.4) into the equation (4.5) and by rearranging the
expression, it will give equation (4.6) and (4.7).
(
⁄ )
(4.6)
(4.7)
At a given influent concentration, organic loading rate (QSi/Vr) and known volume of
anaerobic reactor, equations (4.6) can be used to estimate the concentration of substrate
present in the reactor effluent when KB and Umax values are obtained. Equation (4.7) can be
used to determine the required volume of anaerobic reactor needed to reduce effluent
substrate concentration in order to meet the discharge standard. Equation (4.4) can be used to
determine the KB and Umax of the reactor. The inverse of loading rate [Vr/Q(Si-Se)] can be
73
plotted against the total loading rate of the reactor Vr/QSi. The slope and intercept of the
straight line are KB/Umax and 1/Umax respectively.
4.2.7 Statistical Analysis
Statistical analyses were performed on the measured parameters as well as to test the
differences between the measured and predicted results at an alpha level of 0.05. Graph Pad
Prism v.5, software package was used for statistical analyses and graphs.
4.3 RESULTS AND DISCUSSION
4.3.1 Profiling of Microbial Community Structure of a Full-Scale UASB Reactor Granules
Based on 16S rDNA Analysis
4.3.1.1 Characteristics of granular sludge used for the molecular analysis
The physico-chemical characteristic of granular sludge collected for microbial analysis in this
study were determined using standard methods as described in section 3.2.3 (Table 4.3).
Table 4.3: Characterization of granular sludge
used for molecular analysis
Parameter Concentration (mg/L)
TCOD 1700
SCOD 1220.58
TSS 70.54
VSS 62.27
TS 83.42
VS 70.38
PO₄ 70.59
NO₂ 0.12
NH3 1.5
pH 6.78
Temperature (°C) 28
74
4.3.1.2 Methanogenic Archaea and bacteria detected from the granular sludge using FISH
technique
The preliminary analysis of granular sludge was carried out using FISH technique with
probes targeting Eubacteria and Archaea domains (Table 4.1). In-situ hybridization analysis
of the samples stained with ARC 915 and EUB 388 mix probes revealed the dominance of
both rod and coccoid-shaped methanogens in the reactor (Figure 4.2a-c). Thick cell wall with
long and short curved rods, cocci and irregular cocci packet shapes indicated the presence of
diverse groups of acetoclastic methanogenic Archaea belonging to the order
Methanobacteriales, Methanococcales and Methanomicrobiales. Detection of rod and cocci
packet shapes by ARC915 probe shows that Methanosaeta and Methanosarcinales-like
species are also present in the UASB reactor. The presence of cocci with thick cell wall and
packet-like shape, typical to the genus Methanosarcina was further confirmed by the MS821
probe. Furthermore, the positive hybridization of MX825 probe confirmed the presence and
dominance of acetoclastic Methanosaeta group in the samples (Figure 4.2d-e), which is
distinguished by their typical rod-shape (Raskin et al., 1994; Sekiguchi et al., 2001; Gomec et
al., 2008; Vavilin et al., 2008). These groups of methanogens have previously been reported
to be present in anaerobic reactors which showed more than 70% CH4 production (Krzysztof
and Frac, 2012). The detection is in agreement with the previous findings where the genus
Methanosarcina were detected in granular sludge samples (Sekiguchi et al., 1999;
Jupraputtasri et al., 2005; Kovacik et al., 2010).
75
Figure 4.2: (a) Images of granules hybridized by highly rhodomine labeled archaeal-domain
oligonucleotide probes (ARC915) showing diverse species of methanogens (green) at 1000 x
magnification; (b) corresponding image of ARC915 granules showing diverse species of
methanogens stained with DAPI (blue), (c) granular sludge of FISH labeled with
tetramethylrhodomine-5-isothiocyanate using the universal probes for eubacteria (EUB338),
(d) the MX825 probe labeled sample to confirmed the acetoclastic Methanosaeta group and
(e) the corresponding DAPI stained cells for EUB mix.
a
c d
e
b
10 µm
10 µm
10 µm
76
4.3.2 Community of the Granular Sludge Using PCR
The results of FISH were further confirmed using PCR. The phylogenetic structure of the
bacterial, archaeal and methyl coenzyme-M reductase (mrcA) gene communities was
investigated by 16S rDNA gene-cloning analysis. The full-scale UASB reactor has six
different compartments (bottom to the top; C1, C2, C3, C4, C5, and C6; Figure 4.1), of which
PCR amplicons of both eubacteria and methanogenic Archaea obtained from compartments
1, 3 and 6 were selected for cloning and analysis. The results obtained are discussed in details
below.
4.3.2.1 Bacterial diversity within the reactor compartments
The bacterial populations in the granule samples were analyzed using the domain specific
primer set 27f/1492r that target eubacteria 16S rDNA genes (Giovannoni, 1991). Figure 4.3
shows the bands on agarose gel corresponding to each of the six compartments. The
phylogenetic analysis of the PCR products revealed an abundance of three major bacterial
phyla belonging to the Proteobacteria, Firmicutes and Chloroflexi within the reactor
compartments. The other major phylum detected was an uncultured candidate division WS6
(Table 4.4). Class Gamma and Deltaproteobacterium, Clostridia, Syntrophorhabdus
aromaticivorans and Dehalococcoidetes were also present in abundance in the reactor
samples (Figure 4.4; Table 4.4).
77
Figure 4.3: Agarose gel depicting PCR products for the bacterial fragments (1500 bp). The
bands corresponding to lanes C1–C6 represent the bacterial fragments from the six
compartments of the UASB reactor when PCR amplification was performed using 27f/1492r
specific primer set. Lane L corresponds to the 1 kb DNA marker used in this study.
L C1 C2 C3 C4 C5 C6
1500 bp
250 bp
78
Table 4.4: Bacterial community profiles of the clones retrieved from granular sludge samples
taken from the UASB reactor, as compared to the known sequences in the GenBank database
Source/Habitat/Microorganism Hits Sequence
length
Identity
(%)
Accession
number/Reference
Anaerobic digestion of beet silage 1 1473 95-98 Krakat et al. (2011)
Forest musk deer intestine 6 1451 95 JF690890, JF690880,
JF690878, JF690871,
JF690869, JF690882
Bacterial communities in sediment of
shallow lake Dong ping
1 - 95 Song et al. (2012)
Fiber degrading bacteria from pig faeces - 95 FJ753786, FJ753832
Swine faeces, human faeces,
cellulose/xylan degraded
1 - 98 JX120100, JX006776
Psychrophilic methanogenic community of
wastewater treatment EGSB bioreactor
4 970 96 EU722393, McKeown et al.
(2009)
Granular sludge of full-scale reactor
treating corn straw
13 1470 95 Qiao et al. (2013)
Microbial community composition as
affected by substrate types of anaerobic
digesters
3 1027 92 JX023221
Hydrocarbon and chlorinated-solvent
contaminated aquifer undergoing intrinsic
bioremediation
1 1470-1472 95 Dojka et al. (1998)
Anaerobic swine lagoons 1 1428 95 AY953166
Methane production from hydrocarbon in
oil sand tailings
1366 86 Siddique et al. (2012)
Microbial fuel cells 1 1474 85 Dunaj et al. (2012)
Biological wastewater treatment plant
integrated with constructed wetland for the
treatment of tannery effluent
95 KC110172
Anaerobic digestion of food waste 1506 90 KF699851
Toluene-degrading methanogenic
consortium
1 85 Ficker et al. (1999)
Biogas slurry 1514 95 GU112185
79
Figure 4.4: Phylogenetic tree of bacterial clones obtained from granular sludge of UASB
reactor treating brewery wastewater using universal 27f/1492r bacterial primer set. The
evolutionary history was inferred using the neighbor-joining method (Saitou and Nei, 1987).
The nucleotide sequences were submitted to the National Centre for Biotechnology
Information website under the accession numbers KM065733 – KM065740 corresponding to
the selected clones (1B-10B) from compartments C1, C3 and C6.
80
Detection of four major phyla (Bacteroidetes, Chloroflexi, Firmicutes and Proteobacteria)
with relative differences in the bacterial population in AD systems has been reported
previously (Nelson et al., 2011; Lee et al., 2012; Lee et al., 2014; Jang et al., 2014). A
similar pattern of diverse phylogenetic fingerprint for the bacteria at phylum and genus level
were reported for anaerobic degradation of brewery wastewater, corn straw, as well as birch
and conifer pulp (Werner et al., 2011; Nissilä et al., 2012; Novak et al., 2013; Qiao et al.,
2013).
The clones obtained from compartment 1 showed more than 90% sequence similarity to
uncultured bacterial. Clone B1 from compartment 1 showed 98% similarity with cellulose,
amylase and protease enzyme-producing bacterium P618 in the GenBank as shown in Table
4.4 (JX120100). These enzymes are excreted by hydrolytic and fermentative bacteria during
the hydrolysis stage of anaerobic conversion of complex organic matter in the wastewater
into soluble monomers (Arsova, 2010; Ralph and Dong, 2010; Krzysztof and Frac, 2012).
Clone B8 obtained from compartment 1 was found to be closely related to uncultured
bacterium and uncultured Enterobacteriaceae bacterial clones in the phylum Proteobacteria.
These organisms are involved in the direct production of methanogenic substrates, such as
CO2, H2, formate and acetate. The clones in this compartment also showed 98% sequence
similarity with known sequences of Escherichia ferusonii (NR074902) and Escherichia coli
(JX041515) in the GenBank database.
In compartment 3, the major groups of bacteria were closely related to class
Gammaproteobacteria and uncultured Enterobacteriaceae bacterium (JQ516439). Few
clones were similar to Cronobacter sakazakii (JF690890), formerly known as Enterobacter
sakazakii, a Gram-negative, non-spore–forming, motile and peritrichous rod of the
Enterobacteriaceae family. Furthermore, few other clones showed similarity to uncultured
prokaryote (GU208330) bacteria, uncultured eubacterium WCHB1-06 (AF050595) of the
phylum Firmicutes and class Clostridia and uncultured Dehalogenimonas sp. (JN540166) in
phylum Chloroflexi, toluene-degrading methanogenic consortium bacterium (AF423183)
(96% sequence similarity).
81
Enumeration of Cronobacter sakazakii from sewage sludges has been reported in the
literature (Iversen et al., 2008; Kucerova et al., 2010). The importance of this Cronobacter
sakazakii strain in the treatment of winery effluent using UASB reactor was demonstrated by
Keyser et al. (2003). Its ability to degrade recalcitrant compounds of anaerobically digested
spent wash from an effluent discharge site (Rajasundari and Murugesan, 2011) and its
relevance in the production of H2 as a metabolite that can be used by CH4 producing Archaea
during dark fermentation were mentioned in an earlier study (Kang et al., 2012). Phylum
Firmicutes, genus Clostridium are known to be directly involved in the conversion of
complex organic matter in the industrial waste to the metabolites that can be used directly by
the methanogenic Archaea. They are efficient in the degradation of complex organic matter
and acetic or lactic acid fermentation to CO2 and H2 (Nelson et al., 2011; Wirth et al., 2012).
Other closely related genera of this phylum were observed in compartment 3. Similar
observations were noticed in other studies as reported in the literature (Keyser et al., 2006;
Rincón et al., 2008; Krzysztof and Frac, 2012; Wirth et al., 2012; Sundberg et al., 2013).
In compartment 6, the largest proportion of bacteria belonged to phylum Proteobacteria of
class Delta and Gammaproteobacteria that contain mostly Gram-negative bacteria in their
lineages. Sequence similarity (99%) with known sequences in the GenBank database further
showed that the clones from this compartment belong to class Deltaproteobacteria of family
Syntrophorhabdaceae. They are closely related to the clustered sequence of anaerobic
environmental clones belonging to phylum Deltaproteobacteria (formally known as
Deltaproteobacteria group TA), family Syntrophorhabdaceae and Syntrophorhabdus
aromaticivorans (Figure 4.4).
The abilities of Syntrophorhabdaceae bacteria to digest recalcitrant compounds of spent wash
during anaerobic degradation have been reported (Qiu et al., 2008; Nakasaki et al., 2013;
Shen et al., 2013; Nobu et al., 2014), especially, in brewery wastewater (Werner et al., 2011).
The family Syntrophorhabdaceae contains well-known species of syntrophic substrate-
degrading anaerobes such as those of the genera Syntrophus, Smithella and Syntrophobacter
(Qiu et al., 2008; Nobu et al., 2014). They are known as amino acids degraders and sulphate–
reducing bacteria (SRB) (Shen et al., 2013). Species of the genus Syntrophobacter has the
82
ability to utilize sulphate as an external electron acceptor, but their growth by sulphate
reduction is known to be very slow (Shen et al., 2013; Nobu et al., 2014). Detection of
sulphate–reducing bacteria in this study explained the low to no sulphate in the brewery
effluent (treated wastewater) from the UASB reactor as observed in this study (Figure 4.4).
Raskin et al. (1996) also noticed about 15% of SRB in a methanogenic reactor, even in the
absence of sulphate in the reactors influent. Competition and coexistence of sulphate-
reducing bacteria, acetogens and methanogens in an anaerobic bioreactor was investigated by
Dar et al. (2008). The SRB may be competing with methanogenic organisms for the available
electrons and utilization of acetate particularly at high organic loading rates (Casserly and
Erijman, 2003; Ince et al., 2010).
Furthermore, members of the Syntrophorhabdaceae family isolated from anaerobic treatment
of industrial wastewater have been reported to play an important role in degradation of
aromatic compounds present in the industrial wastewater, especially Syntrophorhabdus
aromaticivorans (Qiu et al., 2008; Nakasaki et al., 2013; Shen et al., 2013; Nobu et al.,
2014). Syntrophorhabdus aromaticivorans is an obligate anaerobic, syntrophic substrate-
degrading mesophilic organism capable of oxidizing p-cresol, phenol, benzoate, isophthalate
and 4-hydroxybenzoate in association with an H2-scavenging methanogen partner
(hydrogenotrophic methanogen) (Shen et al., 2013). The 16S rDNA gene sequence analysis
of clones in compartment 6 were closely related to Syntrophorhabdus aromaticivorans strain
UI of group TA in the class Deltaproteobacteria isolated in granular sludge taken from an
UASB reactor treating manufacturing wastewater (Qiu et al., 2008). Relatively large numbers
of this type of bacteria isolated mainly from methanogenic environments especially UASB
sludge samples have been reported in the literature (Sekiguchi et al., 1998; Wu et al., 2001;
Lykidis et al., 2011).
4.3.2.2 Archaea composition in the granular sludge
The Archaea community, as group of CH4–producing organisms is assumed to be dominant
in the granules obtained from a biogas-producing UASB reactor. Figure 4.5 showed the
Archaea bands on agarose gel corresponding to each of the six compartments using a
universal ARC622f/ARC915r primer set.
83
Figure 4.5: Agarose gel showing 16S rDNA gene PCR fragments obtained from the
amplification of genomic DNA extracted from the granular sludge samples using ARC
primer set. Bands corresponding to lanes C1–C6 represent the Archaea fragments from the
six compartments of the UASB reactor between 243–250 bp using 1 kb DNA marker (lane L)
in the analysis.
Analysis of the clones obtained from the different compartments showed 98-100% sequence
similarity to known sequences of methanogenic Archaea in the GenBank (Table 4.5). The
detected 16S rDNA sequences were affiliated to the Methanobacteriales,
Methanomicrobiales and unclassified archaeon clones. However, members of the
Methanococcales were not detected in the granules using this primer set as shown in the
phylogenetic analysis from the gene sequences obtained from the GenBank database (Table
4.5). This may be due to their growth requirement of high–salt concentration (0.3-0.9% NaCl
(w/v)), which are not normally found in anaerobic reactors (Bialek et al., 2011).
L C1 C2 C3 C4 C5 C6
250 bp
1000 bp
84
Table 4.5: Sequence similarity of Archaea from the full-scale UASB reactor with the GenBank database
sequences
Clone Most closely related
organisms
Accession number Source/Habitant Reference
C1( KM191135) Uncultured archeons
DQ262487, Methanogens from biogas plant Unpublished
HF966604, Methanogens in anaerobic digesters Sousa et al.
(2013)
JN038003,
KC442808, KC352709,
KC182519, AB775722,
AB710147, AB818554
Environmental sample Unpublished
HE648051, HE648045,
HE648044
Anaerobic bioreactors treating
oleate-based wastewater
Salvador et
al. (2013)
Uncultured
Methanobacterium
archeons
JQ247412, JQ247417,
JF754565, JF754562
Ody sludge and its enrichment
amended with alkanes incubated
for over 500 days
Unpublished
Methanobacterium sp.
JF732736
AB288281
Microbial fuel cell
Unpublished
Methanobacterium
formicicum
JN566059, JN243318,
JN205061, JN205059,
JN205052, NR115168
JX042445
Mesophilic corn-fed on-farm
biogas plants and lab scale biogas
fermenters
Methanogen Isolated from the
anaerobic batch reactor of pig
slurry
Unpublished
Unpublished
Methanoculleus sp AB288272 Deep subsurface groundwater from
sedimentary rock
Unpublished
Uncultured
euryarchaeote clone
ANT2-EFL
GU969413
Brazilian Antarctic Station
wastewater
Unpublished
C3 (KM191136)
Methanobacterium
formicicum
HQ591420
Z29436 (DSM 3636)
Anaerobic microorganisms
involved in methanol
transformation in an underground
gas storage facility
Unpublished
Methanobacterium
palustre
NR_041713 Methanogenic rod isolated from a
Philippines rice field
Joulian et al.
(2000)
Uncultured archaeon
clone
HQ438759
AB598266
Soil microcosms contaminated with
phenanthrene
Unpublished
Euryarchaeote clone GU969419 Brazilian Antarctic Station
wastewater
Unpublished
C6 (KM191137)
Uncultured archeons
JN617328
Methanogenic archaeal community
in Lake Taihu
Unpublished
85
In compartment 1, clones obtained from the granular sludge were closely related to phylum
Euryarchaeota, genus Methanobacterium with 98% similarity to Methanobacterium
formicicum (Table 4.5). Genus Methanoculleus sp. (AB288272) of order Methanomicrobiales
was found to be similar to clones obtained from this compartment. This group comprised of ~
85% of the total Archaea clones, while 12% belonged to uncultured archaeon clones (98-99%
similarity). The clones obtained from compartment 3 granular sludge were affiliated to
known DNA sequences of Methanobacteriaceae archaeon, M. formicicum (DSM 3636;
Z29436), Methanobacterium palustre (NR_041713), Methanobacterium sp. clone ARC and
uncultured archaeon (99% sequence similarity) in the GenBank database.
Similar pattern was observed in the clones obtained from the last compartment (C6). The
clones showed 99% similarity with archaeon sp. and Methanobacterium species. The
majority of the clones detected in this compartment were affiliated to the order
Methanobacteriales (CU466652; 99% similarity), family Methanobacteriaceae obtained
from environmental 16S rDNA sequence from Evry wastewater treatment plant anoxic basin
(Chouari et al., 2010), uncultured prokaryote (EU717078; 100% sequence similarity), and
uncultured archeons.
Sequence analysis of the Archaea community showed that closely related species belonging
to the M. formicicum and M. palustre were abundant in the reactor, especially in
compartments 1 and 3. Thus, the presence of this hydrogenotrophic Archaea in sufficient
amounts in the reactor and in the compartments indicated that the rate of H2 conversion
produced by the acetogenic bacteria to CH4 is high in these compartments. This further
confirmed the syntrophic association between acetogenic and the methanogenic bacteria in
the studied UASB reactor (Amani et al., 2011; Ziemiński and Frąc, 2012).
Dominant hydrogenotrophic methanogens of order Methanobacteriales has previously been
reported in a mesophilic reactor (Bialek et al., 2011; Traversi et al., 2011; Zhu et al., 2011;
Salvador et al., 2013). The findings of Leclerc et al. (2004) showed the abundance of
Methanobacteriales among the diverse group of the archaeal community in a UASB reactor
treating brewery wastewater. During brewery wastewater degradation, production of acetate
86
and H2 from ethanol normally occur through the interaction of H2 utilizing Archaea and H2
producing syntrophic bacteria (Ince et al., 2010). As discussed earlier, the UASB reactor
studied was very high in sulphate-reducing bacteria and they are the major competitor with
Methanobacteriales species for H2 in the absence of sulphate (Casserly and Erijman, 2003).
Likewise, Methanobacteriales are the most abundant hydrogenotrophic and acetoclastic
methanogens that were detected in the UASB reactor investigated in this study.
4.3.2.3 Detection of methyl coenzyme-M reductase gene A (mcrA) in the granular sludge
A novel method of metagenomics coupled with FISH is increasingly used to link the genetic
identity of microorganisms to their ecological function in this field (Nercessian et al., 2005).
Characterization of methanogens based on the methanogenic approach using a small subunit
of ribosomal RNA has been used in many studies and their limitations in providing a direct
link to physiology, metabolic capacities, as well as difficulties to determine the functions of
the unknown organism was mentioned (Pycke et al., 2011; Supaphol et al., 2011; Niu et al.,
2013, Buriánková et al., 2013). Identification of methanogens based on 16S rRNA as a
marker is generally limited as methanogens from several different major lines of descent can
only be found within the kingdom Euryarchaeota. The use of a functional marker (mrcA)
genes-based approach encoding the key enzymes of characteristic metabolic pathways that is
exclusive to the methanogenic Archaea to identify the methanogenic population in the
treatment process is well documented (Nercessian et al., 2005; Nettmann et al., 2008; Rastogi
et al., 2008; Krober et al., 2009; Steinberg and Regan, 2009; Traversi et al., 2011; Zhu et al.,
2011; Kampmann et al., 2012; Traversi et al., 2014).
The successfully amplified PCR products using methanogenic specific primers (mcrA) after
sequencing and phylogenetic analysis showed 96 to 100% similarity to methanogenic
Archaea belonging to the order Methanobacteriales and Methanomicrobiales (Figure 4.6).
Similar results were previously reported from the UASB reactor granules treating brewery
wastewater (Diaz et al., 2006) and also from other anaerobic reactors producing biogas
(Castro et al., 2004; Cardinali-Rezende et al., 2009; Kovacik et al., 2010).
87
The mcrA sequences clustered around the Methanobacteriales such as Methanobacterium
beijingense strain, Methanobacterium aarhusense and Methanothermobacter crinale showed
96% sequence similarity (Shlimon et al., 2004; Cheng et al., 2011; Kampmann et al., 2012).
This further confirms the dominance of hydrogenotrophic Methanomicrobiales within the
UASB reactor granules. However, in this study, the amplification of the mcrA primer sets
using PCR did not detect the Methanosarcina and Methanosaeta sp. in the granular samples
as reported by Luton et al. (2002). Most clones belonged to the order Methanomicrobiales
and few clones were Methanosarcina sp., while none was reported for Methanosaeta sp.
(Luton et al., 2002). Similar observations were also made by Castro et al. (2004) and Smith et
al. (2007).
88
Figure 4.6: Phylogenetic tree for methanogenic Archaea obtained from granular sludge of
UASB reactor treating brewery wastewater using methyl coenzyme-M reductase (mcrA) gene
primer set. The evolutionary history was inferred using the neighbor-joining method (Saitou
and Nei, 1987). The GenBank accession numbers are KF715644–KF715648 corresponding to
the selected clones.
HM800561|Uncultured archaeon
HM800578|Uncultured archaeon clone
AB689112|Uncultured archaeon mcrA gene
HM800576| Uncultured archaeon
HM800633|Uncultured archaeon
Clone 2
AB775742|Uncultured archaeon, mcrA gene
AB353235|Uncultured Methanobacteriales archaeon mcrA gene
EF465106|Methanobacterium beijingense strain DSM 15999
Clone1
JF460423|Uncultured archaeon csk_139
Clone 5
AY937274|Uncultured methanogenic archaeon clone GranMCR7M4
AY459317|Uncultured euryarchaeote
DQ260578|Uncultured methanogenic archaeon clone MARMC26
GU447210|Uncultured archaeon
Clone 3
FJ982890|Uncultured methanogenic archaeon
DQ662590|Uncultured euryarchaeote clone MCR-HID-R00-35
EF628128|Uncultured methanogenic archaeon
EF628149|Uncultured methanogenic archaeon
DQ662546|Uncultured euryarchaeote clone MCR-HID-R03-13 tase
EF628183|Uncultured methanogenic archaeon clone CWL-22
AY386125|Methanobacterium aarhusense
HQ714987|Uncultured Methanobacterium sp. clone MW45
JN793940|Uncultured archaeon
EU980413|Uncultured archaeon
EU980409|Uncultured archaeon
Clone 4
AY458405|Uncultured euryarchaeote
JQ686800|Uncultured methanogenic archaeon
JQ686787|Uncultured methanogenic archaeon
FJ226618|Uncultured archaeon
E.Coli
98
82
61
58
71
95
67
95
51 65
36
46
69
5051
89
3617
22
78
53
0.2
89
4.3.3 Optimization of QPCR for Quantification of Microbial Communities Present in the
Granular Sludge Samples
The abundance of Archaea and bacterial 16S rDNA copies was quantified within the different
compartmental of the UASB reactor using the quantitative real-time PCR (QPCR) based
assay. This was done to quantify and compare the microbial populations in the UASB reactor
and to establish the phases of anaerobic fermentation correlation to the microbial data within
the compartments. In contrast to the conventional end-point detection PCR, QPCR
technology has better sensitivity and reproducibility than conventional PCR and can easily be
used in studies requiring a large number of samples (Talbot et al., 2008).
Firstly, the universal primer set for the domain Archaea was tested with the suggested PCR
mixture and thermocycling conditions as modified from the protocol described by Steinberg
and Regan (2009) and this protocol was applied to all primer sets. Due to different amplicon
lengths of the 16S rRNA gene fragment an additional annealing step was included in the
cycling protocols for the ARC and BAC assays to obtain an optimum standard curve.
Known concentrations of standard DNA were used to validate all real-time PCR assays with
determination coefficient (R2) values of 0.991 and 1.000 respectively (Table 4.6). Table 4.6
shows the statistical analysis derived for the constructed standard curves with the
corresponding primer sets. There were no significant differences in the slopes of the standard
curves at 95% confidence interval for each set of primers used regardless of their amplicon
size. This shows the feasibility and accuracy of QPCR assays for the quantification of
microbial 16S rDNA gene copy numbers in the granular sludge samples. This approach has
previously been employed for quantifying microbial DNA from the samples collected from
biogas producing UASB reactors (Brinkman et al., 2003; Hermansson and Lindgren, 2001;
Nadkarni et al., 2002; Suzuki et al., 2000),
The average values of intercept and slope for each primer set were used to quantify the
amount of targeted 16S rDNA copy numbers in the granules (Table 4.6). Average
amplification efficiencies for bacteria and Archaea were 97.6% and 98.8% respectively,
90
which further showed the consistency of the QPCR assay. At the end of each QPCR run,
primer dimer was checked to confirm that there was no non-specific binding during each
reaction using melting curve analysis. The abundance of microbial communities as
determined by QPCR was reported as DNA copy numbers of 16S rDNA genes per nanogram
of genomic DNA isolated from reactor samples.
Table 4.6: Description of QPCR standard curves parameters for 16S rDNA copy number for
ARC as the universal Archaea and BAC as the universal bacterial primer sets that are
responsible for biological conversion of complex organic matter in the brewery wastewater
into simple monomer and CH4 production
Parameter
Primer set
ARC-set BAC-set
Linear range (copies/ng DNA) 1.64 × (104 ~10
11) 2.77 × (10
3 ~10
10)
Slope (standard deviation) -3.565 (0.019) -3.485 (0.011)
R2 of slope 1.000 0.991
Intercept 43.93 41.05
4.3.3.1 Comparison of concentration of Archaea and bacterial communities in the different
reactor compartments
The average 16S rDNA gene copies of Archaea in the samples were calculated against the
total bacterial 16S rDNA gene copies. The compartment showed a noticeable disparity in
terms of the composition of bacteria and methanogenic Archaea population using real-time
PCR (Figure 4.7). It was observed that the concentration of Archaea decreased with an
increase in bacterial concentration along the reactor compartments (1 to 6) as shown in Figure
4.7. There was a correlation between species diversity using PCR and gene copy number
using QPCR. Identification and quantification of the 16Sr DNA using PCR and QPCR
confirmed the variation in the concentration of bacteria and Archaea down the reactor
compartments. There was a reduction in bacteria and an increase in Archaea concentrations at
91
the bottom of the reactor in compartment 1 when compared with compartment 6 using
QPCR. Similarly, the PCR results showed high methanogenic diversity at the bottom of the
reactor (C1) with few known bacteria clones, when compared with identified bacterial and
Archaea species in compartment 6.
Reactor compartments
C1 C2 C3 C4 C5 C6
% o
f to
tal 1
6S r
DN
A p
er n
anog
ram
DN
A
0
20
40
60
80
100
Archaea
Bacteria
Figure 4.7: Variation in the percentage of bacteria and Archaea communities in the granules
collected at the different reactor compartments (C1–C6) using universal primer sets for the
quantitative real-time PCR assay, in this study.
In compartment 1, the percentage of Archaea in the sample was much higher (96.28%) than
the percentage of bacteria (3.78%) (Figure 4.7). However, the percentage of bacterial
increased to 98.34% in compartment 6 with decrease in Archaea percentage (1.66%). There
was a change in the quantity of bacteria along the reactor compartment throughout the
monitoring period.
The results showed variation in the microbial population in each compartment. It can further
be deduced that different compartments in the reactor may have been involved in different
92
phases of anaerobic degradation of organic matter in brewery wastewater with different
concentration of metabolic products been produced as confirmed by the DNA sequencing
results and the QPCR assays. Ye et al. (2009) noticed the abundance of the Archaeal 16S
rDNA gene in the total prokaryotic community quantified from sediment of Lake Taihu using
QPCR. Similar variation in the quantity of archaeal genes along the length of the reactor was
recorded by Kubota et al. (2014) with Archaea colonizing the lower and middle parts of the
reactor as observed in the present study.
Figure 4.8 shows the results of the QPCR assays in terms of gene copy number using the
gDNA samples extracted from the granules obtained from each compartments of the full-
scale UASB reactor using each primer set. The concentration of bacteria as revealed by
QPCR assays showed that bacterial gene copies was dominant and abundant in compartment
C6, but decreases down the reactor compartments (Figure 4.8). Compartment 1 has a
relatively low concentration of bacterial gene copy number, followed by an increase in cell
number in compartment 2 and thereafter, gradual increase in concentration from compartment
3 to 6.
The total concentration of bacteria using QPCR in this study ranged between 2.58 × 103
to
3.43 × 106 copies/ng DNA. The highest concentration of bacterial per nanogram of sample
was observed in samples taken from compartment 6 (3.43 × 106 copies/ ng DNA) and
decreased to 2.58 × 103 in compartment 1. However, fluctuation in the quantity of bacteria in
the different compartments was also noticed. This might be due to competition among the
bacteria for available nutrients or as a result of inhibition of some bacteria through the
activities of other group of bacteria in the reactor or the influence of digestion temperature
(Lee et al., 2012; Welte and Deppenmeier, 2013; Yuan et al., 2014). Apart from the
intermediate metabolites that are produced during the conversion of complex organic matter,
some intracellular material is released when bacteria die which serves as nutrients for other
organisms (Aquino and Stuckey, 2004; Ghosh, 2013).
On the other hand, quantification of 16S rDNA genes using ARC915r/ARC622f revealed that
the proportion of total Archaea varied along the reactor compartments with Archaea
93
colonizing the lower part (C1 and C2) and the middle (C3) of the reactor (Figure 4.8). The
concentration of Archaea decreased from C1 to C6 with higher DNA copies in compartment
2 and lowest concentrations were found in compartment 6. Specifically, on average, the total
concentration of Archaea during the study ranged between 5.80 × 104 to 1.45 × 10
6 copies/ng
DNA. The concentration of Archaea per nanogram of sample was much higher in
compartment 2 (1.45 × 106 copies/ ng DNA) and decreased to 5.80 × 10
4 with an increase in
the reactor‘s compartment (C6). Quantification in terms of percentage of microbial
community showed that the reactor had higher percentage of Archaea in compartment 1 due
to small amount of bacterial concentration in compartment 1; however, compartment 2 had
higher concentration of Archaea and bacteria when compared with compartment 1. The
Archaea domain consists of sensitive organisms; their increase in the lower compartments 1
to 3 also correlates with the presence of low concentrations of toxic substances in those
compartments (Gerardi, 2003; Ali Shah et al., 2014). A reduction in Archaea concentration or
cell number indicated the production of intermediates metabolites that did not favour or
inhibit the growth of methanogens (Botheju and Bakke, 2011). Langenhoff and Stuckey
(2000) also observed a higher methanogenic activity of the bottom part of an anaerobic
reactor treating low strength wastewater.
Thus, a combination of both PCR-based and FISH (RNA-based methods) techniques
produced a better understanding of the microbial consortia present in the UASB reactor
treating brewery wastewater. These techniques helped us to identify and quantify the
microbial population and possible phases at which anaerobic fermentation takes place in the
reactor. This study extends our knowledge on the different hydrolytic, acidogenic, acetogenic
bacteria and methanogenic Archaea present in the granules of the full-scale reactor
investigated.
94
Sample from different UASB reactor compartments
C1 C2 C3 C4 C5 C6
Num
ber
of 1
6S r
DN
A c
opie
s of
m
icro
bial
com
mun
itie
s /n
g D
NA
10x100
100x100
1x103
10x103
100x103
1x106
10x106 Archaea
Bacteria
Figure 4.8: Abundance of Archaea and bacterial DNA copy numbers of 16S rDNA genes per
nanogram of genomic DNA extracted from the granular samples obtained from each
compartments of the full-scale UASB reactor using QPCR assays for the primer sets used in
this study.
4.3.4 Performance of UASB Reactor and Biogas Production
The characteristics of the influent brewery wastewater are shown in Table 4.7. The average
influent COD concentration was 2005.73 ± 1139.85 mg/L at 28˚C (Table 4.7) with a COD
removal efficiency of 78.97 %. The average effluent substrate concentration (Se) from the
UASB reactor was lower (457.25 ± 272.41 mg/L) than the influent substrate concentration
(Si). This might be due to low levels of total solids introduced into the reactor which helped
the reactor performance (Section 3.3.2.2).
95
Table 4.7: Biochemical properties of pre-conditioned
brewery wastewater entering the UASB reactor before
treatment
Parameters Average concentration
values*
Temperature (˚C) 29.21
pH 6.87
COD 2005.73
TSS 2449.46
TS 4520.00
TDS 1792.80
TON 0.52
NH4 21.64
NO2 2.30
NO3 0.07
ORP (mv) -144.78
Sulphate 178.25
Protein content 134.40
Orthophosphates 21.25
Conductivity (mS/cm) 2.18
Alkalinity (mg CaCO3/ L) 2880.52
* All parameters are in mg/L unless otherwise stated.
Table 4.8: Average composition of biogas produced in this study
Biogas composition Values (%)
CH4 65.9
CO2 30.7
N2 3.4
H2S Not Detected
H2 Not Detected
96
The efficiency of COD removal and the methanogenic activity were further shown in the
composition of biogas generated from the UASB reactor with CH4 content of 60-69% (Table
4.8). The ANOVA results showed that CH4 yield depended on the substrate present in the
wastewater in terms of COD removal efficiency as shown in section 3.3.2.4. In addition, the
microbial characterization and biogas production results further confirmed the presence of
methanogens in the UASB sludge.
The influent characterization confirmed the presence of significant amount of VFAs in the
brewery wastewater, which could serve as substrate for the methanogens to produce biogas
(Karakashev et al., 2006). Volatile fatty acids such as acetate (538.30 mg/L), propionic acid
(237.50 mg/L), butyric acid (50.06 mg/L) and valeric acid of 16.54 mg/L were detected in the
influent wastewater with no detection of these acids in the effluent. This shows that the
methanogens metabolized the VFA present in the brewery wastewater to produce CH4.
Among all the methanogens detected using FISH, Methanosaeta sp. and Methanosarcina sp.
were reported to possess the ability to metabolize acetate (Ferry, 1993; Buriánková et al.,
2013). Presence of Methanosarcina in the granular sample can be explained by the acetate
concentration and high biogas production from the reactor (Traversi et al., 2011). The
significance and abundance of Methanosarcina sp. at high acetate level was in agreement
with previous studies (Karakashev et al., 2006; Ariesyady et al., 2007; Rincón et al., 2008;
Vavilin et al., 2008). It is known that members of this genus grow by obligate methyl
reduction with H2 or CO2 reduction with H2 or methylotrophic catabolism of methanol
dimethylsulfide and methylated amines as well as aceticlastic fermentation of acetate
(Maeder et al., 2006; Trzcinski et al., 2010). Delbès et al. (2001) reported that species closely
related to the family Methanobacteriales and Methanobacterium formicicum were found
dominant in an anaerobic bioreactor during acetate accumulation. However, the current study
has shown a reduction in methanogenic activities when there was a high nitrogen and
ammonium-nitrogen content in the effluent. This could be as a result of unfavourable
conditions in the reactor leading to a reduction or inhibition of methanogenic growth in the
reactor. There was almost no nitrites and very low concentrations of nitrates (less than
25mg/L) in the reactor effluent. This shows that nitrate reduction took place in the reactor
because many Archaea and bacteria can utilize nitrate as a source of cellular nitrogen
(Trzcinski et al., 2010).
97
Studies have shown the dynamics and structures of methanogenic populations at various
volatile fatty acid concentrations (Wang et al., 2009) during digestion processes (Yu et al.,
2006). All of these studies discussed the capacity of the microbial communities to respond to
changes in anaerobic environments, such as altered feeding (Kovacik et al., 2010) and
temperature (Sasaki et al., 2011), among others. In addition, the FISH analysis of these
samples have also shown a poor fluorescent signal during hybridization which could be
attributed to a high protein content of the granules due to the low methanogenic activities
(Wikström et al., 2012).
4.3.5 Kinetic Modelling and Model Validation
Kinetic studies are critical for the design and operation of any full–scale reactor to determine
the substrate removal rates. Various kinetic models viz., Monod, Contois, modified Stover-
Kincannon and Grau second order have been tested (Kincannon and Stover, 1982; Vasant and
Barsoum, 2009). Among these, the modified Stover–Kincannon kinetic model was selected
for this present study which has been widely employed for high strength wastewater samples
(Kapdan and Erten, 2007; Turkdogan-Aydinol and Yetilmezsoy, 2010; Yetilmezsoy, 2012).
From equation 4.6, the saturation constant (KB) and the maximum utilization rate constant
Umax in the model was estimated to be 13.64 and 18.51 (g/L/day) respectively. The
application of equation (4.6) by regression analysis showed that the utilization rate was
directly proportional to the reactor efficiency (R2= 0.978; Figure 4.9). The comparison
studies exploring the modified Stover-Kincannon model for anaerobic treatment of different
types of wastewater under different experimental conditions are shown in Table 4.9. From
Table 4.9, the maximum utilization constant (Umax) values (11.83 and 1.996 g/L/day) reported
by Yetilmezsoy (2012) was lower than the value obtained in this study, however, lower than
the estimated value obtained for synthetic-based wastewater (Ahn and Forster, 2000). The
high Umax in the synthetic wastewater could be attributed to the presence of readily
biodegradable substrates that are easily accessible to microorganisms (Ahn and Forster,
2000).
98
Table 4.9: Comparison of different types of anaerobic wastewater treatment processes using the modified Stover-Kincannon model
Digester types
Type of substrates
Operating
temperatures (°C)
Modified Stover-Kincannon model kinetic and estimated coefficients
KB (g/L/day) Umax(g/L/day) R2 References
UASB Brewery wastewater 28-32 13.64 18.51 0.978 Present study
UASB Poultry manure
wastewater
30-34.5 13.02 11.83 0.991 Yetilmezsoy (2012)
Anaerobic
biphasic fixed
film reactor
Distillery wastewater 37 1.69(kg/m3/d) 2 (kg/m
3/d) 0.992 Acharya et al. (2011)
UASB Municipal wastewater 17.1-21 1.536 1.996 0.972 Turkdogan-Aydinol and
Yetilmezsoy (2010)
UASB Synthetic wastewater
(2,4-dichlorophenol)
- 0.0098
(mg/L/day)
0.01 (mg/L/day) 0.992 Sponza and Uluköy (2008)
Anaerobic filter Synthetic wastewater
(saline)
37 5.3 7.05 0.910 Kapdan and Erten (2007)
Mesophilic
anaerobic filter
Synthetic wastewater
(starch)
35 50.6 49.8 0.998 Ahn and Forster (2000)
Mesophilic
anaerobic filter
Paper pulp liquor 35 6.14 6.71 0.998 Ahn and Forster (2000)
99
Figure 4.9: Effect of organic loading rate on COD removal rate using the modified Stover-
Kincannon model to determine the kinetic constants.
Industrial scale wastewaters such as brewery effluent might contain different recalcitrant and
more complex compounds that are less degradable (Yetilmezsoy, 2012). Furthermore, the
operating conditions of anaerobic reactors could also influence the activity of
microorganisms which can affect the kinetic rates. The biochemical and the kinetic data
obtained in this study confirms the efficiency of the microbial community present within the
UASB reactor in degrading the organic matter present in brewery wastewater to produce
optimum biogas that can serve as source of energy. Further, to test the validity of the model,
the observed effluent COD values and predicted values obtained from the model were
compared (Figure 4.10). The results indicated high significance of the model with an
excellent fit between the predicted effluent COD concentrations and the observed
concentrations (P < 0.001) (Figure 4.10). High R2
value of 0.957 between the observed and
predicted values suggested that the predicted results are in accordance with the observed
results (Figure 4.10). This further showed the suitability of the modified Stover-Kincannon
model to predict effluent concentrations from this anaerobic treatment system treating
brewery wastewater.
Slope= 0.737
Intercept = 0.054
R² = 0.978
0
1
2
3
4
5
6
7
8
9
0 2 4 6 8 10 12
V/Q
(Si -
Se)
(g/l
/d)
V/QSi (g/l/d)
100
Figure 4.10: Relationship between the observed and predicted effluent COD concentrations
by modified Stover-Kincannon model.
4.4 CONCLUSIONS
A combination of FISH and PCR techniques helped to identify diverse group of microbial
populations in each compartment. In addition, microbial fingerprinting showed syntrophic
interactions between different bacterial groups and the methanogenic Archaea present in
the reactor.
In-situ hybridization analysis revealed the dominance of methanogenic Archaea of the
orders Methanobacteriales, Methanococcales, Methanomicrobiales and
Methanosarcinales-like species in the granular sludge samples.
Bacterial groups that are required for the decomposition of organic matter in the brewery
wastewater into simple monomers and for production of acetate as the major substrate for
CH4-producing Archaea were detected in this study. The major bacterial communities in
the reactor include the representative from the phyla Proteobacteria, Firmicutes and
Chloroflexi.
R² = 0.957
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.15 0.25 0.35 0.45 0.55
Obse
rved
Se
(g C
OD
/l)
Predicted Se (g COD/l)
𝑆𝑒 = 𝑆𝑖 18.51 𝑆𝑖
13.64 + 𝑄𝑆𝑖
𝑉𝑟⁄
(𝑆𝑒)𝑚𝑜𝑑𝑒𝑙 = 0.862 (𝑆𝑒)observed 0.142
101
Diverse groups of biogas-producing methanogens within the UASB reactor granules
treating brewery wastewater were observed. Methanobacteriales, Methanomicrobiales
and unclassified archaeon clones were detected in the granular sludge taken from the
UASB reactor using PCR. Species detected include Methanobacterium beijingense,
Methanobacterium aarhusense, Methanobacterium formicicum, Methanoculleus sp.,
Methanobacterium palustre and Methanothermobacter crinale using PCR.
The modified Stover–Kincannon model was found to be applicable to predict effluent
COD concentrations from the anaerobic reactor treating brewery wastewater.
It is hoped that the characterization of eubacteria and methanogenic Archaea in the
granules used for this study will bridge the gap of knowledge on the microbial ecology of
the UASB reactor investigated. This will further help engineers to apply appropriate
operational and environmental conditions that will select appropriate microbial
community for efficient reactor performance.
4.5 RESEARCH OUTPUTS
a) Journal articles
1. Enitan, A. M., Kumari, S., Swalaha, F.M., Adeyemo, J., Ramdhani, N. and Bux, F.
(2014). Kinetic modelling and characterization of microbial community present in a full-
scale UASB reactor treating brewery effluent. Microbial Ecology, 67:358–368.
b) Conference Papers
1. Enitan, A. M., Kumari, S., Swalaha, F.M. and Bux, F. Real-time PCR for quantification
of methanogenic Archaea in a UASB reactor treating brewery wastewater (2014)r.
Conference of the International Journal of Arts & Sciences, CD-ROM. ISSN: 1943-
6114: 07(03):103–106.
2. Enitan, A. M., Kumari, S., Swalaha, F.M. and Bux, F. Use of mcrA-targeted real-time
quantitative PCR for quantification of methanogenic communities in reactor treating
brewery wastewater. Presented at Water Institute of Southern Africa (WISA) Conference,
Mbombela, Mpumalanga, South Africa, 25-29 May, 2014 (Oral presentation).
102
CHAPTER FIVE: DEVELOPMENT OF A MATHEMATICAL MODEL TO
DESCRIBE THE BEHAVIOUR AND PERFORMANCE OF A UASB
REACTOR TREATING BREWERY WASTEWATER FOR BIOGAS
PRODUCTION
5.1 INTRODUCTION
Recovery of bioenergy from spent biomass, industrial wastewaters and other types of waste is
commonly achieved through the conventional anaerobic digestion (AD) process (Demirel et
al., 2010). AD technology, such as the upflow anaerobic sludge blanket (UASB) reactor
technology is used for the treatment of different types of wastewaters for biogas production.
The efficient functioning of biogas production systems provides different benefits to users
and the community, resulting in energy and cost savings, environmental protection and
conservation of resources (Tiwari et al., 2006; Rajput et al., 2012). However, bioconversion
of organic substances to biogas depends on many operational factors (Oktem and Tufekei,
2006). Sometimes reactors may fail or encounter serious problems, depending on factors such
as influent composition, pH, temperature, OLR, HRT and carbon to nitrogen ratio of the
source material. These factors also affect microorganisms that are responsible for the
degradation of organic matter in the bioreactors (Senturk et al., 2013).
A UASB reactor depends on granular sludge as the core unit in order to convert the organic
component of wastewater to biogas (Batstone et al., 2002; Liu et al., 2003). The sludge
granules consist of dense microbial communities that typically include various bacterial
communities in the sludge bed (Enitan et al., 2014a) and the gas-liquid-solid phase at the top
of the UASB reactor helps in sludge retention. Optimal operational conditions such as HRT,
upflow velocity, influent COD, pH and temperature are needed for efficient biological
treatment of wastewater to produce biogas in the UASB reactor (Wiegant, 2001). Thus, it is
important to improve the operational parameters in order to enhance the efficiency of the
UASB digestion process particularly for the production of methane (CH4)-rich biogas. This
could be done by several methods such as predicting and optimizing the operational
conditions; satisfying the nutritional requirements of microbes by using different biological
and chemical additives and manipulating the feed proportions (Yadvikaa et al., 2004). Some
other ways include the recirculation of digested slurry, returning microbes back into the
reactor and modifying existing biogas plant design (Yadvikaa et al., 2004). Hence, an in-
103
depth understanding of process dynamics including the (i) feedstock characteristics, (ii)
operational and environmental parameters, (iii) reactor design and (iv) the microbial ecology
are important for the optimization of AD systems.
A simple mathematical model that describes some of the conditions that define the anaerobic
treatment process is a generally accepted approach in defining the specific parameters of
system performance. Models based on process kinetics can be used to understand the
underlying biological and transport mechanisms within the reactor (Acharya et al., 2011)
thus, giving more useful information on the state of the reactor and any impending failure.
Recently, mathematical modelling of bioreactors has greatly helped in controlling and
improving the treatment efficiency of such systems, as well as in facilitating the experimental
procedure to enhance the degradation of organic material in the waste feedstock used for
biogas production (especially CH4) (Blumensaat and Keller, 2002; Jeong et al., 2005; Lübken
et al., 2007; Mu et al., 2008; Zhou et al., 2011) and to improve the effluent quality (Acharya
et al., 2008). Models have been used to account for reactor performance along with the
associated principles and conditions that affect CH4 production (Reungsang et al., 2012). In
addition, models could be used to predict the compounds that are produced or consumed as
well as the rate of production (Nadais et al., 2011; Thorin et al., 2012). The results of
modelling can be used to estimate treatment efficiencies and system characteristics of full–
scale reactors operating under similar conditions.
To study the kinetics of biogas formation from complex organic matter, two approaches can
be adopted. The first approach is to find the rate-limiting substrate for the kinetic evaluation
and the second approach is the use of COD or volatile solids concentration as an indicator of
substrate concentration (Chen and Hashimoto, 1978). Methane production is said to be
directly related to COD removal, and biogas yield is not the same as CH4 yield because the
composition of biogas comprises of CH4, CO2, water vapour, and a few other gases, such as
hydrogen sulphide and hydrogen gas (Krishna, 2013).
104
Several studies have been carried out on the development of suitable models that best explain
the conditions that will enhance the conversion of organic substances present in the
wastewaters to biogas (CH4) production during AD (Batstone et al., 2000; Batstone et al.,
2002; Colussi et al., 2012; Parsamehr, 2012). However, one of the main drawbacks of the
available mathematical models for anaerobic reactors is their complexity. Several models
based on different concepts and parameters have been reported to be difficult to apply to a
UASB reactor, because they involve many variables (Zhao et al., 2010; Colussi et al., 2012;
Thorin et al., 2012). The application of these models is limited by the parameters needed to
describe them.
For this reason, the development of an applicable model for a UASB reactor with the aim of
reducing the complexity will be helpful for better understanding of the behaviour of the
reactor and to enhancing bioenergy generation. More studies are needed to derive simple and
convenient models that can predict and optimize biogas yield, especially CH4. This paper
presents a model that describes the kinetics of an intermittent-flow UASB reactor treating
brewery wastewater based on mass balance principles. We considered that untreated COD as
the primary substrate with no additional oxidizing agents added into the reactor would be
converted to biogas (CH4 and CO2) (Zainol, 2012). We considered the reduction of COD to
hydrogen gas and hydrogen sulphide insignificant in this study (Chen and Hashimoto, 1978).
At standard temperature and pressure (STP), the digestion of 1 g COD added is equal to the
formation of 0.35 L of CH4. Thus, knowing the influent COD concentration and quantity, we
could deduce the volume of CH4 produced from a reactor. The remaining COD in the reactor
could then be calculated and the energy equivalent released through AD of the wastewater
could be determined, because most of the energy contained in biogas is represented by CH4.
Thus,the developed model describes the behaviour of the reactor with respect to substrate
degradation and the effect of endogenous decay rate on CH4 production based on modified
Chen-Hashimoto equations by Ghaly et al. (2000).
105
5.2 MATERIALS AND METHODS
5.2.1 Ghaly et al. (2000) Model
Various mathematical models have been proposed to describe substrate and biomass
concentrations as well as biogas production in a batch, or continuous process reactor (Colussi
et al., 2012; Fdez-Güelfo et al., 2012; Zainol, 2012). Among these models for AD, Ghaly et
al. model (2000) was found to be suitable forthis study. The governing equations for the
process are obtained from the mass balance of substrate and concentration of biomass in the
reactor compartment. The model follows Monod kinetics. The principle of the process is
based on modified Chen-Hashimoto equations, in which the concentration of biomass in the
system depends on the growth and decay rate of microorganisms under steady–state
conditions for an intermittent flow of organic matter into the biological treatment unit.
5.2.1.1 The microbial mass balance
The microbial mass balance of an UASB reactor (Figure 5.1) was described as follows by
Ghaly et al. (2000):
Microbial change rate = Microbial input rate + Microbial growth rate - Microbial death rate-
Microbial output rate (5.1)
The microbial growth rates in a batch experiment have traditionally been measured, in which
a single species of microorganisms passes through a logarithmic growth phase during the
conversion of the organic substrate. The microbial growth rate, dX/dt, is described by;
(5.2)
which can be written as;
. (5.3)
106
Figure 5.1: Schematic diagram of a single compartment of an upflow anaerobic sludge
blanket reactor (See abbreviations for definition of symbols).
During steady-state conditions, the biomass concentration in the influent is negligible (Xi
0), compared to the biomass concentration in the reactor. In addition, Xr is equal to Xe due to
perfect mixing in a completely mixed reactor. The rate of substrate removal from the reactor
is therefore neglected. In steady-state conditions, dX/dt = 0 and equation (5.3) can be
rearranged to obtain equation (5.4). Thus,
( )
Q = V (µ – Kd) (5.4)
Equation (5.4) can be rewritten as;
(5.5)
The hydraulic retention time, θh, is defined as V/Q. The inverse of θh can be substituted into
equation (5.4) as;
µ – Kd =
. (5.6)
As shown in equation (5.6), the net specific growth rate is µ – Kd.
Xe, Q, Se
Effluent
Reactor
Xi, Q, Si
Influent Vr, Xr, Sr
107
5.2.1.2 Substrate mass balance and effluent substrate concentration
The rate of substrate balance in the UASB reactor can be expressed using equation (5.7) and
mathematically as equation (5.8) (Ghaly et al., 2000):
[Substrate change rate] = [Substrate input rate] – [Substrate utilization rate] – [Substrate output rate]
(5.7)
Mathematically, equation (5.7) can be written as;
( )
. (5.8)
At steady state, equation (5.8) was divided by V, and Q/V was substituted for θh. At
equilibrium the substrate balance of a working system was obtained as;
( )
. (5.9)
Thus, under perfect mixing of the reactor content (Xr = Xe), the microbial mass concentration
in the effluent can be written as equation (5.10). This gives the concentration of
microorganism in the effluent as;
( )
( ), (5.10)
Where, (Si –Se)/θh is the rate of substrate utilization. Contois (1959) defined the relationship
between limiting substrate concentration and specific growth rate for effluent substrate
concentration as;
. (5.11)
Under perfect mixing (Se = Sr and Xe = Xr), the association between the rate–limiting
substrate concentration and specific growth rate can be expressed as;
. (5.12)
108
Equations derived from the combination and rearrangement of equations (5.6), (5.10) and
(5.12) are;
( ) (5.13)
( ) , (5.14)
Where, equation (5.14) shows that the influent substrate concentration is inversely
proportional to the substrate concentration in the final effluent.
5.2.1.3 Biogas production
In the reactor, the biodegradable COD is proportional to (Bo–B). Bo is directly proportional
to the biodegradable COD loading rate (Zainol, 2012). Therefore, from equation (5.14), the
CH4 yield (B) can be described by;
( ) . (5.15)
Methane production per gram of substrate (COD) added, B can be described by;
[
( ) ]. (5.16)
Since B is equal to the volume of CH4 produced per unit of COD added, the volumetric CH4
production rate, Yv is equal to B, multiplied by the organic loading rate, Si/θh. The equations
describing the theoretical CH4 output rate per unit of reactor volume therefore, are written as
equations (5.17) and (5.18):
(5.17)
[
( ) ] . (5.18)
109
5.2.2 Modified Methane Generation Model (MMGM)
Ghaly et al. (2000) model does not consider temperature or the amount of non-biodegradable
COD of the feedstock, which are important factors in wastewater treatment. We now describe
a new Modified Methane Generation Model (MMGM), which integrates the effect of
temperature and non-biodegradable COD with the model described above (Ghaly et al.,
2000; Equation 5.18) for a UASB reactor under anaerobic conditions. The assumptions made
include the following:
The UASB reactor was treated as a single compartment.
It was considered as a completely mixed system with continuous influent flow into the
reactor and no return of microbial solids back into the reactor (it is non-recycling).
The substrate was a single biodegradable substance.
Substrate consumers were uniformly distributed in the reactor (bed and blanket) under
perfect mixing.
Reactor operation is at steady state.
The kinetic model follows first–order kinetics using the Monod model with respect to
substrate and biomass concentration.
The developed model‘s outcomes include the quantification of the growth rate of biomass,
substrate consumption and the effect of endogenous decay on biogas formation. The ultimate
CH4 yield coefficient Bo is assumed to be constant based on the literature survey. Studies
have shown that Bo depends on the OLR, sludge or HRT used during the treatment of
brewery wastewater (Oktem and Tufekei, 2006; Fdez-Güelfo et al., 2012). Oktem et al.
(2006) investigated a pilot–scale UASB reactor for the treatment of brewery wastewater in
the mesophilic range. An increase in CH4 yield of 0.25–0.30 m3CH4/kgCODremoved was
observed when OLR was increased with a rise in COD removal efficiency from 60% to 95%.
Similar observation was reported by Chen and Hashimoto (1978) and Yetilmezsoy (2012), on
the value of Bo. The authors mentioned that the value of Bo depends on the type of waste that
is being treated and environmental conditions. Most especially, bioreactor temperature was
mentioned to affect the ultimate CH4 yield coefficient (Chen & Hashimoto, 1978;
110
Yetilmezsoy, 2012) hence, we added operational temperature to our equation (5.18). Chen
and Hashimoto (1978) defined an empirical relationship between the maximum specific
microbial growth rate (µm) and temperature (T) for temperatures between 20˚C and 60˚C on
the analysis of a data set obtained from the literature as equation (5.19) (Yetilmezsoy, 2012).
µmax = 0.013 (T) – 0.129 (5.19)
Studies have shown that maximum specific microbial growth rate in the Chen and Hashimoto
equation (5.19) depends on operational temperature and it increases linearly as the
temperature increases (Yetilmezsoy, 2008; Turkdogan-Aydinol et al., 2010). Therefore,
equation (5.19) can be substituted into equation (5.18) to obtain equation (5.20).
[
[ ( ( ) )]
( )
] . (5.20)
According to equation (5.20), the theoretical CH4 output for any given values of Si and θh is
determined by the specific characteristics of the biodegradation of substrate and the kinetic
constants (µmax and K). In addition, the value of K, according to the Monod equation, may be
associated with the ability of microorganisms to degrade the substrate present in the waste to
produce CH4. A high K value is an indication that the microorganisms present in the reactor
have greater difficulty in converting the organic matter to CH4 (Fdez-Güelfo et al., 2012).
The physicochemical parameters such as temperature have been shown to be the primary
factors affecting µmax. The effect of temperature on µmax could be described by the empirical
relationship mentioned in equation (5.19); for K, the concentration of the organic matter in
the substrate and for Bo the kind of substrate. The biodegradable substrate in the reactor in
terms of its COD concentration is considered to be directly proportional to the actual CH4
generated under normal conditions of temperature and pressure and the fraction of non-
biodegradable COD was included in the model. The fraction of the non-biodegradable COD
111
(nbCOD) was written as equation (5.21) with respect to the initial substrate concentration and
P, as the fraction of the biodegradable COD removed:
nbCOD = (1 – P) 0 ≤ P ≤ 1 (5.21)
Hence, equation (5.20) can be written as shown below, which indicates that the biodegradable
substrate concentration in the reactor is directly proportional to the actual CH4 volume. Then,
the governing equation (5.22) for modified methane generation model (MMGM) can be
obtained as:
( )
[
[ ( ( ) )]
( )
] . (5.22)
The kinetic constant K shows the level of microbial growth in the digestion process. This is
an extension of Ghaly et. al. (2000) model. This model can be used for anaerobic processes
at steady–state operation under perfect mixing and also takes into consideration the material
balance for a mixed reaction; the substrate being the rate-limiting factor. The design and
operation of an AD system is based on fundamental knowledge of kinetics and stoichiometry
of biological reactions. Thus, prediction of industrial–scale anaerobic reactor performance
based on UASB technology in treating brewery wastewater depends on the estimated values
of parameters such as K, µmax, Kd , Y and Bo. However, the kinetic values estimated from
laboratory–scale data are inadequate to describe the actual plant performance (Sykes, 1995;
Iqbal and Guria, 2009). Thus, it is important to determine these parameters from the actual
full–scale treatment plant data, such as the influent and effluent COD concentration, VSS
concentration in the reactor, flow rate and reactor volume. The determination of model
coefficients (K, Bo, µmax, and Kd) is important for the validation of the model, to predict and
to optimize not only the volumetric CH4 production rate of any UASB reactor treating
brewery wastewater, but other different wastewater sources.
112
5.2.3 Determination of MMGM Parameters (K, µmax, Kd , Y and Bo)
The determination of a first-order reaction is represented by Chen and Hashimoto (1978).
They developed a kinetic model based on substrate utilization of the Contois model as;
(5.23)
This model has been widely adopted and used in many studies in the investigation of
anaerobic treatment of high strength wastewater (Cecchi et al., 1992; Yetilmezsoy, 2012;
Zainol, 2012). Equation (5.23) becomes equation (5.24) when divided by µmax.
(5.24)
In a completely mixed system,
and
(5.25)
Let,
(5.26)
Then, the first–order kinetic constant coefficients K and µmax can be determined by plotting
θh against S using equation (5.25). The ultimate CH4 yield (Bo) can be determined using a
least–squares method through nonlinear regression of 1/θh versus CH4 yield. The endogenous
decay constant, Kd can be determined as a function of HRT and VSS values using equation
proposed by Bhunia and Ghangrekar (2008), equation (5.27) or (5.28). These equations can
be used to obtain the values of Kd by plotting a linear regression of 1/θh against (Si – Se)/
(Xeθh). The intercept is equal to Kd and Y is the slope of the straight line that passes through
the plotted points.
113
( )
(5.27)
Or
( )
(5.28)
5.2.4 Software Used and Statistical Analysis
Data obtained from the full-scale reactor were used to derive the parameters in the developed
model. Nonlinear and linear regressions were fitted to data using the GraphPad Prism v5.0
program as the statistical software. Nonlinear regression was conducted based on a least-
squares method to analyzed the predicted CH4 yield and volumetric CH4 production rate.
Correlation using the coefficient of determination between the observed and the predicted
production values was carried out, the probability of fit was calculated and accepted when p
<0.05. The MMGM governing equation (5.22) was coded and simulated using the MATLAB
7.14 software (R2010a, The MathWorks, Inc. Natick, Massachusetts, USA).
5.2.5 Description of the UASB Reactor System Used and Wastewater Sampling
An industrial full–scale UASB reactor treating brewery wastewater was used as described in
section 3.2.1. The biogas produced in the reactor was separated from the effluent and the
biomass in three-phase separators at the top of the reactor was collected in a gas holder
(Tedlar bag, Sigma-Aldrich) for analysis (Section 3.2.3.1). A series of pre-screened brewery
wastewater (reactor influent) and the full–scale UASB reactor effluent ready to discharge into
the municipal sewer system were collected in one–litre sterile glass bottles and transported to
the laboratory at 4°C. Physico-chemical analyses were conducted within 48 hours of
collection with the necessary preservation techniques adapted from Standard Methods
(APHA–AWWA–WPCF, 1998). Physico-chemical tests were carried out as mentioned in
section 3.2.3. Tests were carried out in duplicate.
114
5.2.6 Calculation of Methane Potential and Yield (United Nations Economic Commission
for Europe, 2004)
( ) ( ) ( )
(5.29)
( ) ( ) ( ) (5.30)
(5.31)
5.3 RESULTS AND DISCUSSION
5.3.1 Estimated MMGM Parameters
Removal efficiencies for both BOD5 and COD were found to be ~80% and 79% respectively
and the mean biogas (CH4 content) produced was 65.9%. This indicated that the organic
matter in the industrial wastewater was converted to usable biogas with good effluent
composition. Figure 5.2 shows the time–course for the performance of full–scale UASB
reactor in treating brewery wastewater during the monitoring process (Section 3.2.2), in–
terms of COD and BOD5 removal efficiencies over the time period.
115
Figure 5.2: The time–course of COD and BOD5 removal efficiencies for the full–scale
UASB reactor treating brewery wastewater in this study.
Table 5.1 shows the experimental data used to determine MMGM parameters as shown in
Table 5.2. The first-order kinetic coefficients K and µmax were determined by plotting θh
against S using equation (5.25) (Figure 5.3). The graph produced a straight line with µmax
given by 1/intercept and K as slope/intercept. The values of µmax and K derived in this study
were 0.117 dˉ1
and 0.046 g/g, respectively (Table 5.3). The kinetic parameters could then be
used to determine the behaviour of a system or bioreactor, which would help to characterize
the microbial-substrate interaction for better treatment efficiency. The ultimate CH4 yield
(Bo) was determined using a least–squares method through the nonlinear regression of 1/θh
and CH4 yield.
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12
Rem
oval
eff
icie
ncy
(%
)
Duration (Weeks)
BOD removal
COD removal
116
Table 5.1: Average data obtained from the full-scale UASB reactor treating brewery
wastewater
θh (h) Q (L/h) COD loading
rate (g/L)
Si (g/L) Se(g/L) Xe (g/L) CH4 production
(L/h)
CH4 yield (L/g
COD added)
8 167 171.43 1.03 0.51 0 224.30 0.18
9 180 167.24 0.93 0.23 2.19 44.35 0.27
9 300 929.53 3.10 1.01 4.40 219.11 0.24
11 180 520.05 2.89 0.43 1.00 154.67 0.30
12 250 248.66 1.00 0.23 6.11 66.88 0.27
12.1 156 170.16 1.10 0.11 4.00 53.70 0.32
13 300 900.62 3.00 0.23 1.73 291.34 0.32
Table 5.2: Data used for the determination of MMGM parameters
θh 1/θh Xe Xeθh Si Se Si-Se Xeθh/(Si-Se) (Si-Se)/(Xeθh) S=(Si-Se/Se)
8
9
0.13
0.11
0
2.19
0
19.71
1.03
0.93
0.51
0.23
0.52
0.70
0
28.00
0
0.04
1.00
3.13
9 0.11 4.40 39.60 3.10 1.01 2.09 18.98 0.05 2.06
11 0.09 1.00 11.03 2.89 0.43 2.46 4.49 0.22 5.66
12 0.08 6.11 73.34 1.00 0.23 0.76 95.96 0.01 3.32
12.1 0.08 4.00 48.40 1.10 0.11 0.98 49.21 0.02 9.18
13 0.08 1.73 22.52 3.00 0.23 2.78 8.12 0.12 12.20
Table 5.3: Estimated MMGM parameters as obtained using the data collected from the
full–scale UASB reactor treating brewery wastewater
Parameter Estimated value Units R2 P-value
µmax 0.117 dˉ¹ 0.709 0.017
K 0.046 g/g 0.709 0.017
Kd 0.083 dˉ¹ 0.767 0.009
Bo 0.516 L CH4/g COD added 0.988 0.006
Y 0.357 g/g 0.7670 0.0088
117
Figure 5.3: Estimation of the kinetic parameter K and the maximum growth rate of
microorganism‘s µmax, from data collected from the full–scale UASB reactor treating brewery
wastewater. The plot of θh against S [where, S = (Si–Se/Se)] gives a straight line with
1/intercept as µmax and slope/intercept as K.
Figure 5.4 shows the graph of CH4 yield against 1/θh with the intercept, Bo corresponding to
0.516 L CH4/g CODadded. The estimated endogenous decay coefficient, Kd value is
represented by the intercept of the straight line graph shown in Figure 5.5 as 0.083 d-¹, while
the slope Y, corresponds to 0.357 g/g. The estimated model coefficients for brewery
wastewater used in this study are shown in Table 5.3. The values are within the range of
values reported in the literature for mesophilic AD for waste types that include wastewater,
banana stem and peel waste, palm oil mill wastewater, dairy manure and the organic fraction
of municipal solid waste from a full-scale plant (Table 5.4). The value of µmax obtained from
the full-scale UASB reactor treating brewery wastewater was higher than the value (0.111 d–
1) reported by Zainol (2012) and lower than 0.135 d
–1 reported by Fdez-Güelfo et al. (2012).
However, the value of Bo is very similar to those reported in the literature (Table 5.4). Hence,
the values of coefficients K, Bo, µmax and Kd were used to validate the model and to predict
treatment efficiency, determine the HRT for treatment of wastewater, and predict volumetric
CH4 productivity of an UASB reactor treating brewery wastewater.
R² = 0.709
6
8
10
12
14
16
0 2 4 6 8 10 12 14
Hydra
uli
c re
tenti
on t
ime
(θh)
S = (Si–Se/Se)
118
Figure 5.4: Ultimate methane yield (Bo) obtained from data collected from the full–scale
UASB reactor treating brewery wastewater by plotting methane yield against the reciprocal
of hydraulic retention time.
Figure 5.5: The endogenous decay coefficient, Kd and the growth yield coefficient, Y were
calculated from the intercept and slope of the straight line of the plotted graph using the data
obtained from the full–scale UASB reactor treating brewery wastewater.
Bₒ= 0.516
R² = 0.988
0
0.1
0.2
0.3
0.4
0.5
0.6
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
CH
4 yie
ld (
L/g
CO
Dad
ded
)
Reciprocal of hydraulic retention time, 1/θh (d-1)
Y = 0.357 d˗1
Kd = 0.083 g/g
R² = 0.767
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0 0.05 0.1 0.15
1/θ
h
(Si–Se)/(Xeθh)
119
5.3.2 Validation of the Modified Methane Generation Model
The values of K, µmax, Kd, Bo and θh presented in Table 5.3 were used in the model to
simulate methane yield. The simulations were carried out for a fixed substrate concentration
at different hydraulic retention times based on equation 5.16. The simulation indicated
methane yield as a function of hydraulic retention time. The application of the model was
shown by regression analysis of the predicted methane yield with determination coefficient of
0.991 at 95% confidence range with P value of 0.0001. Only 0.009% of the total variations
could not be explained by the regression analysis. A high coefficient of determination R2 of
Table 5.4: Kinetic parameters obtained in this study compared to other studies
Substrate Bo (L CH4 /g
CODadded)
K (g/g
CODadded)
µmax
(dˉ¹)
Kd
(dˉ¹)
Reference
Brewery wastewater 0.516 0.046 0.117 0.083 This study
Banana stem waste 0.326 0.33 0.111 - Zainol (2012)
Synthetic organic fraction of
municipal solid waste
1.167ᵃ - 0.238 - Fdez-Güelfo et al. (2012)
Organic fraction of municipal
solid waste from a full-scale
composting plant
1.15ᵃ - 0.135 - Fdez-Güelfo et al. (2012)
Distillery spent wash - - 2ᵇ - Acharya et al. (2011)
Vegetable product—pea 0.36 - - - Maya-Altamira et al. (2008)
Vegetable product—leek &
fried onion
0.36 - - - Maya-Altamira et al. (2008)
Banana peel 0.277ᶜ - 0.089 - Gunaseelan (2007)
Palm oil mill wastewater 0.381 - 0.304 - Faisal and Unno (2001)
Dairy manure at 25°C 0.230ᶜ 0.883ᶜ 0.279 0.038 Ghaly et al. (2000)
Dairy manure at 35°C 0.230ᶜ 0.883ᶜ 0.317 0.036 Ghaly et al. (2000)
Brewery wastewater - - 0.022 0.037 Anderson et al. (1996)
a = l methane /g DOC ; b = Kg mˉ³dᶜ = l methane/g VS added
*All abbreviations are in abbreviation section.
120
0.991 shows a strong Goodness of fit of the model. Figure 5.6 showed the expected behaviour
when compared with experimental values obtained from the full-scale reactor investigated.
There was a strong correlation coefficient of 0.747 between the predicted and the observed
values for methane yield, which showed the applicability of the model to predict methane
yield.
.
7 8 9 10 11 12 13 14
0.0
0.2
0.4
0.6Predicted
Observed
R = 0.747
Hydraulic retention time h (h)
Met
hane
yie
ld (
L/g
CO
Da
dd
ed)
Figure 5.6: Observed and predicted methane yields at different hydraulic retention times.
In order to further validate this model, the observed volumetric methane production rate and
predicted values obtained from MMGM were compared at different temperatures and OLR.
For a randomly selected operating scenario, a volumetric organic loading rate between 2.0
and 11.8 g COD/L/day and the initial substrate concentration, Si = 6 to 12 g COD/L, (Bo =
0.516, T = 26°C and 32°C) using MMGM (Equation 5.22) showed that increasing the
volumetric OLR to 8.26 g COD/L/day would stimulate the methane yield better
(corresponding to the maximum volumetric methane production rate of Yv = 1.46 L CH4/g
CODadded/day).
121
In order to evaluate the fitness of MMGM, the predicted values of the volumetric methane
production rates were plotted against the observed values for different organic loading rates
(Figure 5.7a) and when the OLR increased from 2.0 to 8.26 g COD/L/day, the predicted Yv
increased from 0.29 to 1.46 L CH4/g CODadded/day. However, Yv decreased as the OLR rose
to 11.80 g COD/L/day. The coefficient of determination value (R2 = 0.994) for the methane
production rate showed the goodness of fit of the developed model (MMGM). The coefficient
of determination (R2 = 0.994) showed that 99.4% of the variance in the model can be
explained by the model and the model was shown to be extremely significant with p <
0.0001.
A similar trend was noticed in the observed methane production rates although; there was
fluctuation in the observed values due to operational and environmental parameters.
However, the highest value for the observed Yv (0.51-0.83 L CH4/g CODadded/day) was
recorded at OLR between 4.4 to 9.29 g COD/L/day, and the observed Yv decreased when the
OLR reached 11.80 g COD/L/day. The data indicates that the volumetric methane production
rate fluctuate with an increase in OLR, hence values higher than 0.8 g COD/L/day were not
included in the relationship shown in Figure 5.7b. Up to this point, the correlation between
the predicted and the observed Yv was very strong (R2 = 0.990), showing a linear relationship
between these parameters at different OLR (Figure 5.7b). A noticeable decrease in Yv as
observed at higher organic loading rates suggested that OLR could influence the kinetic
parameters due to the presence or accumulation of inhibitors or toxic compounds in the
reactor and also reduce volatile solids removal, thus affecting the volumetric methane
production rate (Babaee and Shayegan, 2011). However, at higher OLRs the values between
observed and predicted methane production rate vary considerably and the MMGM
overestimated the methane production rates.
122
0.0 0.5 1.0 1.5 2.0
2.10
2.16
2.60
2.76
3.08
4.44
8.26
9.29
11.85
Observed
Predicted
(a)
Methane production rate, Yv
(L methane/ g CODadded/d)
Org
anic
lo
adin
g r
ate (
g C
OD
/L
/day
)
0.0 0.2 0.4 0.6 0.80.0
0.2
0.4
0.6
0.8
y = 0.8256x - 0.0565 (R2 = 0.990)
(b)
Observed, Yv
Pre
dict
ed, Y
v
Figure 5.7: (a) The trend between observed and predicted volumetric methane production
rates at different organic loading rates using the newly developed model and (b) the scatter
plot of predicted vs observed volumetric methane production rates at lower organic loading
rates.
123
A similar trend was reported in the AD of banana stem waste (Zainol, 2012) and a UASB
reactor treating poultry manure wastewater (Yetilmezsoy, 2012). However, overloading the
bioreactor with a high substrate concentration has been reported to be one of the factors
contributing to the reduction in the methane production rate. A reduced methane production
rate signifies the presence of a possible inhibiting factor in the process, such as a decrease in
pH as a result of an increase in the concentrations of VFAs (Tiwari et al., 2006; Yetilmezsoy,
2012).
The influence of HRT and OLR on the microbial communities and the performance of an
anaerobic reactor to treat olive waste at steady state have also been investigated (Rincón et
al., 2008). The authors observed the maximum methane production rate of 1.7 L CH4 STP/L
day when the OLR was increased from 1.50 to 9.29 g COD/L/day at 17 days HRT. However,
when the OLR was increased to 11.0 g COD/L/day at HRT of 15 days, there was a reduction
in the pH value (from 7.5 to 5.3) as well as increase in the effluent total VFA by about 400%
(Rincón et al., 2008). This further confirmed that the OLR affects the value of methane
production rate.
The effect of operational temperature on the activity, survival and growth of the microbial
consortium in an AD system was reported by Khalid et al. (2011). The effect of operational
temperature (26–32°C) on the volumetric methane production rate (Yv) was simulated using
the developed model (equation 5.22). The predicted volumetric methane production rate at
29°C was higher than that at other temperatures (Figure 5.8). The regression analysis showed
the goodness of fit of the developed model with strong determination coefficient of 0.862 and
the adjusted determination coefficient of 0.882. This confirms the applicability of the
modified methane generation model to predict volumetric methane production from a UASB
reactor treating brewery wastewater.
Several studies have shown the crucial effects of even a slight change in the operating
temperature on biogas production, especially its CH4 content. Any sudden change might lead
124
to a drastic decrease in biogas production due to change in microbial populations and reduced
CH4 content and volume (Chae et al., 2007; Ward et al., 2008). Chae et al. (2007) reported
the maximum CH4 yield at 35°C when compared to that at 30°C and 25°C. Therefore, for
better treatment efficiency and high volumetric methane production rate, operating
temperature should be optimized for the reactor design and operation (Ward et al., 2008;
Yetilmezsoy, 2012).
Observed methane production rate Yv (L methane /gCODadded/day)
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Pred
icte
d m
etha
ne p
rodu
ctio
n ra
te Y
v (L
met
hane
/gC
OD
adde
d/day
)
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Observed Yv vs Predicted Y
v
Figure 5.8: The predicted and observed volumetric methane production rates at different
temperatures using the developed model (MMGM).
5.4 CONCLUSIONS
We developed a modified methane generation model (MMGM) for an UASB reactor
that treats brewery wastewater and validated it with respect to substrate degradation
and the effect of endogenous decay rate on the methane production.
Quantification of model parameters indicated that the composition of the wastewater
strongly affects the kinetics of the digestion process.
125
The developed model (MMGM) predicted the rate of methane production for AD of
brewery wastewater at different temperatures and OLRs.
There was a strong correlation between the predicted and the observed measured
values for methane production rate. The predicted results further showed that a change
in operational conditions (OLR, influent substrate concentration, the HRT and
operational temperature) could significantly affect methane production rate.
The model is easy to use due to its simplicity with only a few variables that facilitate
the calibration of the model. It is believed that this model could be used to predict
methane production rate of anaerobic digestion process treating brewery wastewater.
5.5 RESEARCH OUTPUT
1) Abimbola M. Enitan and Josiah Adeyemo. Estimation of Bio-kinetic Coefficients for
Treatment of Brewery Wastewater. Oral presentation at the World Academy of Science,
Engineering and Technology Conference, New York, USA, June 5-6, 2014. International
Science Index, 8(6): 365-369.
126
CHAPTER SIX: MULTI-OBJECTIVE OPTIMIZATION OF A METHANE–
PRODUCING UASB REACTOR USING A COMBINED PARETO MULTI-
OBJECTIVE DIFFERENTIAL EVOLUTION ALGORITHM
6.1 INTRODUCTION
Environmental pollution, especially water and air pollution, has become a challenging task
for both engineers and scientists in the world. Currently, research has shifted to biofuels as
alternative renewable sources due to depletion of fossil fuels. Biofuels produced by AD of
organic materials in industrial wastes through the synergistic metabolic activities of microbial
consortia include biogas (Gueguim Kana et al., 2012). In several European countries, AD is
employed to treat more than 10% of organic matter present in the industrial wastes thereby,
saving chemicals (Gueguim Kana et al., 2012). However, the industrial viability of this
process requires a suitable combination of chemical and physical process parameters and a
low-cost substrate; hence there is a need for process optimization for efficient system
performance to produce sufficient biofuel.
There are many optimization problems in science and engineering that require maximization
of system desirable properties and simultaneously minimizing its undesirable characteristics.
A significant portion of research and applications in the field of AD optimization has focused
on single–objective optimization problems, whereas most of the natural world problems
involve multiple-objectives which are conflicting in nature (Babu et al., 2005; Iqbal and
Guria, 2009; Kusiak et al., 2009; Abu Qdais et al., 2010). Multi–objective optimization
problem (MOOP) involves finding one or more optimum solutions to more than one
objective optimization problem (Deb, 2001). The aim of MOOPs is to simultaneously
optimize a set of conflicting objectives to obtain a group of alternative trade-off solutions
called Pareto-optimal or non-inferior solutions which must be considered equivalent in the
absence of specialized information concerning the relative importance of the objectives
(Adeyemo and Otieno, 2010; Deb, 2011).
Currently, optimization problems are represented as an intelligent search problem, where one
or more agents are employed to determine the optimal on a search landscape, representing the
constrained surface for the optimization problem (Das et al., 2008). A large portion of control
127
problems exhibits multiple stage and multiple objective (MSMO) characteristics. Likewise,
AD processes also involve several decision making branches resulting in many objective
functions and constraints. Despite this prevalence, there are few methods with the capability
to solve general large-scale conflicting multi-objective optimization problems.
Evolutionary algorithms (EAs) are computational-based biological–inspired optimization
algorithms. They are stochastic searching methods, commonly used for solving non-
differentiable, non-continuous and multimodal optimization problems based on Darwin‘s
natural selection principle (Enitan and Adeyemo, 2011; Sendrescu, 2013). Evolutionary
algorithms are widely used for single and multi-objective optimization in AD processes in
relation to methane production (Babu et al., 2005; Iqbal and Guri, 2009; Wei and Kusiak,
2012).
Evolutionary algorithms use several variables of a problem to provide an optimum solution.
Evolutionary algorithms can generate Pareto optimal solutions for different AD models with
equally good solutions with respect to all objectives; none of the solutions should dominate
another (Deb, 2001; Enitan and Adeyemo, 2011). Studies have shown that EAs are good
alternative methods for monitoring state variables in biotechnological processes (Babu et al.,
2005; Soons et al., 2008; Iqbal and Guria, 2009).
Some of the most frequently used evolutionary multi-objective optimization algorithms for
AD include non–dominated sorting genetic algorithm (NSGA), multi-objective genetic
algorithm (MOGA), multi-objective differential evolution algorithm (MDEA), multiobjective
differential evolution (MODE) and multi-objective particle swarm optimization (MOPSO)
(Srinivas and Deb, 1994; Babu et al., 2005; Adeyemo and Otieno, 2010; Wei and Kusiak,
2012).
Successful applications of DE to batch fermentation process, optimization of non-linear
chemical processes, optimization of process synthesis and design problems, optimization of
biomass pyrolysis and optimal design of shell and tube heat exchangers have been reported in
the literature (Babu and Chaurasia, 2003; Babu et al., 2005; Angira and Babu, 2006). Among
128
other improved versions of differential evolution that have been reported in the literature
include hybrid differential evolution (HDE) (Tsai and Wang, 2005), Pareto differential
evolution approach (PDEA) (Madavan, 2002), MDEA (Adeyemo and Otieno, 2009b), multi-
objective differential evolution algorithm (MODEA) (Ali et al., 2012) and more recently, a
Combined Pareto Multi–Objective Differential evolution (CPMDE) algorithm (Olofintoye et
al., 2014).
Olofintoye and coworkers (2014) developed a combined Pareto multi-objective differential
evolution algorithm for solving multi-objective optimization problems. The CPMDE
algorithm has strength in multi-modal function optimization as demonstrated by Adeyemo et
al. (2014). The algorithm combines methods of Pareto ranking and Pareto dominance
selections to implement a novel selection scheme at each generation. The algorithm employs
harmonic average crowding distance measure as against NSGA that implements a crowding
distance. The superiority of harmonic average crowding distance has been demonstrated by
Huang et al. (2005).
The CPMDE algorithm has been successfully applied to various engineering problems
(Olofintoye et al., 2014; Adeyemo et al., 2014), where the ability of CPMDE in solving
unconstrained, constrained and real-world optimization problems was also demonstrated.
Their simulation results show that the CPMDE approach can generate a better Pareto-front
for the selected problems.
The main aim of this chapter is to optimize a methane producing UASB reactor using a
CPMDE algorithm and provide parameter settings for operating the reactor for more efficient
methane generation and better effluent quality. It will be interesting to investigate if the
algorithm will perform better using real-life optimization problems such as anaerobic
treatment of wastewater for better and more robust solutions for the decision makers. In
recent times, a slightly similar problem was solved using different industrial wastewater and
algorithm, but we have now used an improved algorithm (CPMDE) to solve the problem to
have better solutions.
129
The modified methane generation and adopted Stover-Kincannon kinetic models (Enitan et
al., 2014a) were used for the optimization. This is the first application of a CPMDE algorithm
in the area of anaerobic treatment. It is also the first reported multi-objective optimization
study on a brewery wastewater treatment plant for methane production, effluent COD
reduction and biomass concentration.
6.2 METHODS
6.2.1 Optimization of UASB Reactor
The optimization problem was formulated for multi-objective optimization problem of an
existing plant that has been scaled-down for easy optimization and simulation in pilot-scale
reactor. The optimization problem was formulated for maximization of methane production
rate (Yv; Equation (6.1)), minimization of effluent biomass concentration (Xe; Equation (6.2))
and effluent COD concentration (Se; Equation (6.1)). The constrained optimization problem is
written as;
Maximize f1 (P, θh, Si, T) = Yv (6.1)
Minimize f2 (Si,Q) = Se (6.2)
Minimize f3 (Si, θh, Q, Se) = Xe (6.3)
Model equations
( )
[
( ( ) )
( )
] (6.4)
( )
(6.5)
(
⁄ )
(6.6)
130
The decision variables were bounded as;
Si,L ≤ Si ≤ SiU
(6.7)
QL ≤ Q ≤ QU (6.8)
θh,L ≤ θh ≤ θhU
(6.9)
PL ≤ P ≤ PU
(6.10)
TL ≤ T≤ TU
(6.11)
Subject to constraints X ≤ Xe (6.12)
S ≤ Se (6.13)
V = Vr (6.14)
Where, Yv is the volumetric methane production rate, θh is the mean hydraulic retention time,
Si and Se are the influent and effluent COD concentration respectively, while P is the COD
removal efficiency. Xe is the concentration of biomass in the discharge effluent (biomass
wash-out) and OLR is the organic loading rate. Q represents the influent flow rate of
wastewater; T is the operational temperature, while V is the desired reactor volume.
Equation (6.4) is the governing equation to optimize volumetric methane production rate in a
given reactor volume (Vr) in the multi-objective optimization problem formulated. The
important decision variables and inequality constraints are shown in Table 6.1. In this problem,
plant treatment efficiency depends on the biomass concentration in the reactor. Therefore,
prevention of sludge or biomass washout from the reactor is needed for effective treatment
and to meet the environmental discharge requirements, as well as increasing the methane
production rate for biofuel. In equation (6.5), the desired value for biomass wash out from the
reactor was considered as 0.025 g/L. The desired reactor volume of the existing plant is 1400
m3, but for easy optimization and simulation in the pilot-scale reactor, it was scaled–down to
35 m3. The results of the optimization can then be scaled-up to the actual volume of the large-
scale UASB reactor.
131
The discharge of low effluent COD concentration to meet the standard limits is another
important factor for environmental monitoring, as well as the production of substantial
amount of biogas that is rich in methane. Therefore, it was logical to get an optimum
operating condition that minimized the effluent discharge COD for any OLR, Q and Si. Thus,
the minimization of effluent substrate concentration using the modified Stover-Kincannon
kinetic model (equation (6.6)) was included in the optimization. In equation (6.6), the desired
effluent COD concentration was considered as 0.05 g/L.
The boundary conditions for the decision variables based on the scaled-down industrial
process are shown in Table 6.1. The lower and upper limits on θh were decided based on the
HRT of the industrial treatment plant. The microbial consortia in the treatment plant have
been found to be sensitive to temperature changes (Akarsubasi et al., 2006; Krakat et al.,
2010; Khemkhao et al., 2012), which in turn can affects the rate of methane production.
Therefore, operating temperature should be considered as one of the important factors. In this
regard, the minimum and maximum values of temperature were selected based on the
operating range of the industrial plant.
Table 6.1: Details of model-based multi-objective
optimization problem studied using CPMDE algorithm
Objective function Problem
First Maximixe Yv
Second Minimize Se
Third Minimize Xe
Inequality Constraints
Vr (L) = 35
Se (g/L) ≤ 0.05
Xe (g/L) ≤ 0.025
Bounds
Si (g/L) 1 ≤ Si ≤ 10
Q (L/day) 1 ≤ Q ≤ 20
θh (h) 1 ≤ θh ≤ 12
P 0.8 ≤ P ≤ 1
T (˚C) 10 ≤ T ≤ 35
132
The lower and upper limits for the influent substrate concentration were set based on the
capacity of the treatment plant. The minimum and maximum values for the efficiency of
substrate utilization of the reactor at the end of the treatment period in terms of COD removal
were considered. This was to ensure maximum conversion of organic matter to methane and
good effluent quality in order to meet the discharge standard. The lower and upper limits for
influent flow rate were chosen based on the industrial activities and wastewater that the
industry is producing; however the volume in this study was scaled-down to 35 m3 for a pilot-
scale reactor.
6.2.2 Combined Pareto Multi-Objective Differential Evolution (CPMDE) Algorithm
6.2.2.1 The CPMDE algorithm
In this study, a combined Pareto multi-objective differential evolution (CPMDE) algorithm
was used to optimize the formulated mathematical models. The algorithm combines methods
of Pareto ranking and Pareto dominance selections to implement a novel selection scheme at
each generation (Olofintoye et al., 2014). At each iteration of the CPMDE, the combined
population of trial and target solutions is checked for non-dominated solutions. Solutions that
will proceed to the next generation are selected using a combined Pareto ranking and Pareto
dominance selection scheme (Mezura-Montes et al., 2008). After generating a trial
population, tournaments are played between trial solutions and their counterparts in the target
population at the same index. Diversity among solutions in the obtained non-dominated set is
promoted using a harmonic average crowding distance measure (Huang et al., 2005;
Olofintoye et al., 2014) to select the solution that will proceed to the next generation, if
solutions are feasible and non-dominated with respect to each other.
In the CPMDE, boundary constraints are handled using the bounce-back strategy (Price et al.,
2005). This strategy replaces a vector that has exceeded one or more of its bounds by a valid
vector that satisfies all boundary constraints. In contrasts to random re-initialization, the
bounce-back strategy takes the progress towards the optimum into account by selecting a
parameter value that lies between the base vector parameter value and the bound being
violated (Babu et al., 2005). Equality and inequality constraints are handled using the
constrained-domination technique suggested by Deb (2001). The DE/rand/1/bin variant of
133
DE is used as the base for CPMDE. The CPMDE algorithm is summarized as follows
(Olofintoye et al., 2014):
1. Input the required DE parameters such as number of individuals in the population
(Np), mutation scale factor (F), crossover probability (Cr), maximum number of
iterations/generations (gMax), number of objective functions (k), number of decision
variables/parameters (D), upper and lower bounds of each variable, etc.
2. Initialize all solution vectors randomly within the limits of the variable bounds.
3. Set the generation counter, g =0
4. Generate a trial population of size Np using DE‘s mutation and crossover operations
(Price et al., 2005)
5. Perform a domination check on the combined trial and target population and mark all
non-dominated solutions as ―non-dominated‖ while marking others as ―dominated‖.
6. Play domination tournament at each population index.
i. If the trial solution is marked ―non-dominated‖ and the target is marked
―dominated‖ then the trial vector replaces the target vector.
ii. If the trial solution is marked ―dominated‖ and the target is marked ―non-
dominated‖ then the trial vector is discarded.
iii. If both solutions are marked ―dominated‖, then replace the target vector if it is
dominated by the trial vector or if they are non-dominated with respect to each
other.
iv. If both vectors are marked ―non-dominated‖, then note down the index and
proceed to the next index. When all solutions marked ―non-dominated‖ from
steps i – iii above are installed in the next generation, then sort out all solutions
noted in step iv one at a time using the harmonic average crowding distance
measure (Huang et al. 2005). The solution with a greater harmonic average
distance is selected to proceed to the next generation.
7. Increase the generation counter, g, by 1. i.e. g = g+1.
8. If g < gMax, then go to step 4 above else go to step 9
9. Remove the dominated solutions in the last generation
10. Output the non-dominated solutions.
*Note domination checks are performed using the naive and slow method suggested by (Deb,
2001).
Source: (Olofintoye et al., 2014).
134
Olofintoye et al. (2014) evaluated the performance of CPMDE using common difficult test
problems obtained from multi-objective evolutionary computation literature. The ability of
the algorithm in solving unconstrained, constrained and real optimization problems was
demonstrated and competitive results obtained from its application suggested that it is a good
alternative for solving multi-objective optimization problems. Furthermore, based on an
argument by Deb (2001) that most of these test problems are not tuneable and it is difficult to
establish the feature of an algorithm that has been tested, the CPMDE has further been tested
using on tuneable multi-objective test problems (Adeyemo et al., 2014). CPMDE has been
applied to solve real world multi-objective problems and results obtained corroborate the
efficacy of CPMDE in solving multi-objective optimization problems.
6.2.2.2 Implementation of CPMDE algorithm for optimization of an UASB reactor
The ability of CPMDE in solving unconstrained, constrained and real-world AD optimization
problems is demonstrated herein. The principle of CPMDE algorithm includes coding of the
models, decision variables, the constraints as well as evaluation of the fitness function and
improvement of the fitness function using differential evolution operators such as tournament
selection, crossover and the harmonic average crowding distance measure. The crossover
constant, (Cr) and the mutation scaling factor, (F) were set at 0.1 and 0.9 respectively.
Population size, Np was set to 50 and the algorithm was run for a maximum number of
generations, gMax from 300-5000 on different optimization problems. Harmonic average
crowding distances were computed using two nearest neighbours. Further details on the
implementation of CPMDE may be found elsewhere (Adeyemo et al., 2014; Olofintoye et
al., 2014).
6.3 RESULTS AND DISCUSSION
The kinetic model for methane production rate, effluent substrate COD and biomass
concentration were simultaneously optimized in this study to obtain global optimal solutions
from the conversion of organic matter in the brewery wastewater. The kinetic coefficients for
the model equations used are summarized in Table 6.1. These models were optimized by
using the CPMDE algorithm on a computer with dual core processor and 8GB RAM
135
processor. The model equations were first coded and tested with MATLAB software to
ensure that the codes were free of error (Chapter Four and Five). Subsequently, CPMDE
algorithm was used to solve the models as multi-objective optimization problem.
A multi–objective optimization problem involving three-objective functions was solved
simultaneously using the CPMDE algorithm. These include (i) maximization of volumetric
methane production rate, (ii) minimization of effluent discharge COD and (iii) minimization
of biomass wash-out from the treatment plant. For this problem, the constraints and decision
variables used are shown in Table 6.1. The best value of CPMDE optimization parameters for
the three–objective functions are shown in Table 6.2.
Cr- crossover constant, F- the mutation scaling factor
Figure 6.1 shows the Pareto optimal solutions for these three-objective functions. Equally
good solutions with regard to all objectives were obtained for this problem; none of the
solutions dominated another. It was found that as the volumetric methane production rate
increased (improved), both the effluent discharge COD and biomass wash-out from the
treatment plant also increased (worsens) over the entire Pareto optimal surface (Deb, 2001;
Table 6.2: The CPMDE parameters used for multi-objective optimization problem
Parameters Value
Number of Vectors: 50
Number of Parameters: 5
Number of DE generations: 5000
DE control parameters: Cr F
Value 0.1- 0.9 0.1- 0.9
Step 0.1 0.1
Optimization
Number of objectives: 3
Number of constraints: 4
Number of nearest neigbours: 2
Number of non-dominated solutions in final current population 50
Computational time, min 3.17
136
Liu and Wang, 2008; Iqbal and Guria, 2009; Enitan and Adeyemo, 2011). From these results,
none of the solutions dominated any other. All the solutions on the Pareto front were found to
be equally good and were expected to provide flexibility for the solutions on the Pareto front.
Each point on the Pareto optimal front corresponds to a set of decision variables as shown in
Table 6.1. Some of the advantages of using these three-objective optimization problem
include to have a wide choice of solutions and operating points in the Pareto set, because
each point on the Pareto set is obtained from a set of decision variables.
Figure 6.1: Pareto optimal set of solutions obtained for the simultaneous optimization of
volumetric methane production rate (Yv), effluent biomass concentration (Xe) and effluent
substrate concentration (Se) as a multi–objective optimization problem.
The decision variables were further plotted against volumetric methane production rate and
effluent biomass concentration to determine the conflicting variables (Figure 6.2a-d).
However, we noticed a nearly constant decision variables (T, Si and Q) over the range of
Pareto set, thus the results were not plotted. In addition, the degree of scatter of θh and P were
found to be slightly higher with unsmooth Pareto front for the simultaneous optimization of
these objective functions. A similar result was reported by Babu et al. (2005) when MODE
0.0225
0.0230
0.0235
0.0240
0.0245
0.0250
0.0255
0.032
0.034
0.036
0.038
0.040
4.96
4.98
5.00
5.02
5.04
Xe (
g/L
)
S e (g
/L)
Yv (L CH
4 /gCOD/day)
0.0225
0.0230
0.0235
0.0240
0.0245
0.0250
0.0255
137
and NSGA algorithms were employed for solving multi–objective optimization problems of
industrial adiabatic styrene reactor. Iqbal and Guria (2009) explained that scattered optimal
values of the decision variables compensate for each other due to additional objective
function and decision variable to the optimization problem.
However, CPMDE is able to give a more uniform distribution of solutions, than those
reported by Yee et al. (2003) and Babu et al.(2005) using NSGA and MODE respectively.
Furthermore, several other studies have been reported to have encountered scattered decision
variables (Sareen and Gupta, 1995; Tarafder et al., 2005; Khosla et al., 2007). Better spread
shows that CPMDE algorithm found more operating policies that were not discovered by any
other algorithms from which the decision maker could choose from. That is, we have more
options for operating the reactor to produce more methane during anaerobic degradation of
industrial wastewater.
In addition, the methane production rate is observed to increase due to an increase in
hydraulic retention time. This suggested that the higher the time the wastewater spent in the
reactor, the higher the gas production in the reactor. The optimal values of methane
production rate take the upper bound at different θh and high substrate removal efficiency. At
higher effluent flow rate (Q = 14 L/day, Vr = 35), optimal θh took almost the lower limit
between 8-9 h, and increase in Yv was observed as θh decreased. In Figure 6.2(c), it was noted
that the Xe decreased with increase in HRT as the COD removal efficiency (P) remained high
(Figure 6.2d). It may be deduced from the optimal results that high P value between 85-87%
and 8-9 h HRT at 30-31˚C were responsible for the low and almost constant effluent substrate
and biomass concentration with a high methane production rate. This suggested that at high
influent substrate concentrations and flow rate, high COD removal efficiency and Yv
depended on the time the wastewater spent in an anaerobic reactor. The results further
showed that the decision variables at mesophilic temperature are responsible for the scattered
Pareto solutions in the optimized problems for the three–objective functions as shown in
Figure 6.2(a-c). Hence, the simulation models could be used to check the operational
parameters for getting the best effluent quality, biomass washout and the highest methane
production rate in the UASB reactor for the treatment of brewery wastewater.
138
Volumetric methane production rate, Yv (L CH
4/ gCOD/day)
4.92 4.94 4.96 4.98 5.00 5.02 5.04 5.06 5.08
Hyd
rau
lic r
ete
ntio
n tim
e,
h (
h)
9.0
9.5
10.0
10.5
11.0
11.5
(a)
Volumetric methane production rate, Yv (L CH
4/ g COD/day)
4.92 4.94 4.96 4.98 5.00 5.02 5.04 5.06 5.08
CO
Dre
mo
va
l effic
ien
cy, P
(%
)
83.5
84.0
84.5
85.0
85.5
86.0
(b)
Effluent biomass concentration, Xe (g/L)
0.0220 0.0225 0.0230 0.0235 0.0240 0.0245 0.0250 0.0255
Hyd
rau
lic r
ete
ntio
n tim
e,
h (h
)
9.0
9.5
10.0
10.5
11.0
11.5
12.0(c)
Effluent biomass concentration, Xe (g/L)
0.0220 0.0225 0.0230 0.0235 0.0240 0.0245 0.0250 0.0255
CO
D r
em
ova
l effic
ien
cy, P
(%
)
83.4
83.6
83.8
84.0
84.2
84.4
84.6
84.8
85.0
(d)
Figure 6.2: The Optimal decision variables (a) θh and (b) P plotted against volumetric
methane production rate (Yv), as well as (c) θh and (d) P plotted against effluent biomass
concentration (Xe) for the optimized problem.
139
Consequently, the multi–objective optimization conditions within the framework of objective
functions based on the holistic kinetic models using the CPMDE algorithm demonstrated a
useful instrument for simultaneous optimization of various operational parameters needed for
successful running of an UASB reactor. The strength of the integrated multi-objective
optimization approach in this study can be applied to for large-scale applications (from pilot–
to full–scale reactor). It should be noted that the holistic approach presented in this study is
restricted by some boundary conditions and assumptions. However, it can readily be used as a
preliminary analysis before transferring the initial concepts to the full–scale reactor, since the
model coefficients are obtained from the data collected from the full–scale reactor. Based on
the present optimization study, a set of optimal operating conditions was obtained which can
enhance the plant performance without affecting the plant configuration. With regards to
these facts, future works can consider scaling-up the results obtained in this study to the full–
scale system.
6.4 CONCLUSIONS
In this study, optimization of industrial wastewater treatment plant was carried out using
combined Pareto multi-objective differential evolution algorithm.
Modified methane generation and the Stover–Kincannon kinetic models were used for the
optimization of anaerobic reactor treating brewery wastewater for better effluent and
methane production.
A multi–objective optimization problem was solved in this study using CPMDE
algorithm as the optimization tool in order to determine the overall optimal operating
conditions of anaerobic reactor treating brewery wastewater.
The associated objective functions were: (i) the maximization of volumetric methane
production rate, (ii) minimization of discharge effluent substrate concentration and (iii)
the minimization of biomass washout from the reactor. Pareto–optimal sets of equally
good non-dominated solutions were obtained for the multi–objective optimization
problem considered. The decision variables followed the same trend that further proved
the reliability of the results obtained in this study. It also showed that the objectives can
further be improved. However, it is difficult to compare the results obtained in this study
140
with other studies in the literature due to different substrate and decision variables
involved, as well as the algorithm used.
This study is the first application of using combined Pareto multi–objective differential
evolution algorithm for AD optimization of brewery wastewater for better methane
production and effluent quality.
The simulation results showed that the CPMDE algorithm can generate a better Pareto-
front for the selected problem. Its ability to solve unconstrained, constrained and real-
world optimization problem was also demonstrated.
This will benefit the existing reactors and the design of new reactors treating brewery
wastewater, in order to use an optimum environmental condition that will favour the
growth of desired microorganisms to desirable end-products.
The optimization method presented in this chapter has been found to be quite general and
flexible to improve the reliability of design and performance of an existing anaerobic
treatment plant or a new plant. It can be applied to an UASB reactor to enhance its
robustness and performance for better discharge effluent quality and biogas production
with high methane content.
6.5 RESEARCH OUTPUT
(a) Book Chapter
Enitan, A.M., Adeyemo, J., Bux F. and Swalaha F. M. 2014. Multi-objective optimization
of a methane-producing UASB reactor using a combined Pareto multi-objective differential
evolution algorithm. EVOLVE - A Bridge between Probability, Set Oriented Numerics, and
Evolutionary Computation V. Advances in Intelligent Systems and Computing, Springer,
288: 321-334.
141
CHAPTER SEVEN: GENERAL CONCLUSIONS AND
RECOMMENDATIONS
In summary, the composition of raw brewery wastewater obtained from beer producing
industry in KwaZulu-Natal, South Africa was characterized and the efficiency of the full–
scale UASB reactor treating the wastewater was monitored over a period of one year. The
microbial diversity of the granular sludge samples obtained from the UASB reactor were
analyzed using latest molecular techniques with the domain and group-specific rRNA-
targeted oligonucleotide probes and primers. The identification of microbial community
structure was carried out using FISH and PCR techniques, while QPCR was employed for the
quantification of 16S rDNA gene copy numbers in the given samples. A modified methane
generation model (MMGM) in terms of kinetics of an intermittent-flow UASB reactor to
convert brewery wastewater to biogas with high methane content on the basis of mass
balance principles was developed with respect to substrate degradation and the effect of
endogenous decay rate on the CH4 production. In addition, a modified Stover–Kincannon
kinetic model was adopted to predict the final effluent quality and a model-based multi-
objective optimization was carried out using a CPMDE algorithm with set of constraints and
decision variables for the overall optimization of the UASB reactor.
The raw wastewater from the brewery industry was found to be very high in organic matter,
nutrients and solids content which does not meet the required effluent regulatory standards,
however, it was suitable for microbial degradation with pH adjustment. The performance of
the on-site full-scale UASB reactor that treats the above-mentioned raw wastewater was
monitored and the results showed the efficiency of the reactor to reduce the concentration of
organic matter to a permissible level for discharge. However, there is a need to improve the
performance of the reactor in terms of biogas production (methane content), as well as
reducing the ammonia and orthophosphate concentration of the final effluent (after
treatment). The effect of an increase in VFA concentration as a result of decrease in pH was
observed to have a negative effect on methane concentration and reactor‘s efficiency. The pH
of the reactor effluent was within the optimal range for anaerobic bacteria (6.6 and 7.3) at
mesophilic temperatures with a 12 h HRT. Volatile fatty acids were detected in the influent
wastewater with no detection of these acids in the effluent. These acids served as substrates
142
for methanogens to produce biogas during anaerobic degradation of complex organic matter
in the brewery wastewater.
The preliminary analysis of the granular sludge samples using fluorescence in-situ
hybridization technique employing domain specific and group specific probes (ARC 915 and
EUB 388 mix) revealed the dominance of both rod and coccoid-shaped methanogens and
eubacteria in the reactor. Diverse group of methanogenic Archaea belonging to the order
Methanobacteriales, Methanococcales and Methanomicrobiales, as well as Methanosaeta
and Methanosarcinal-like species were detected using ARC 915 and MX825 probes.
The study of the different compartments of the full-scale reactor using PCR analysis
demonstrated a substantial variation and changes in microbial populations. Symbiotic
relationships between the bacteria that are involved in conversion of complex organic matter
to simple monomers were observed using 16S rDNA sequence analysis. The major bacterial
phyla belonging to Proteobacteria, Firmicutes and Chloroflexi needed to convert complex
organic matter in the brewery wastewater to the simple metabolites required by methane-
producing Archaea were detected in the analysed compartments. The sequence obtained
showed (99%) similarity to Enterobacteriaceae bacterium clone, Cronobacter sakazakii,
uncultured Dehalogenimonas sp., uncultured Syntrophorhabdaceae bacterium and
Syntrophorhabdus aromaticivorans.
The microbial fingerprint of the functional gene (mcrA) and the universal Archaea primer sets
revealed the diversity of the methanogenic populations in the granular sludge samples using
PCR analysis. All samples from the different compartments showed positive results for the
primer sets used and produced PCR products of high number of cells with the mcrA genes.
The clones after sequencing and analysis displayed similarity (>97%) to the order
Methanomicrobiales, Methanobacteriales and Methanosarcinales belonging to
hydrogenotrophic and aceticlastic methanogens. Species detected include Methanobacterium
beijingense, Methanobacterium aarhusense, Methanobacterium formicicum, Methanoculleus
sp., Methanobacterium palustre and Methanothermobacter crinale.
143
To understand the distribution and activity of eubacteria, methanogenic Archaea within the
different compartments of the reactor (Figure 4.8), a quantitative real-time PCR (QPCR)
approach was employed. The results of QPCR assays revealed that the bacteria copy number
was dominant and abundant at the upper part (C6) of the reactor (Figure 4.8), and decreased
down the reactor compartments (C1). On the other hand, quantification of Archaea 16S
rDNA copy number revealed that the proportion of total Archaea varied along the reactor
compartments with Archaea colonizing the lower and middle part of the reactor. Lower
concentrations of methanogenic Archaea were observed in compartment 6 using the domain
specific primer set. In this study, variation in the microbial populations at various levels with
the depth of the UASB reactor using PCR and QPCR suggested that each compartment is
responsible for different phases during anaerobic fermentation of organic matter presence in
the brewery wastewater. We could thus conclude that hydrolytic to methanogenic organisms
are present in the reactor and the four stages of AD process do occur at the different
compartment of the full-scale UASB reactor investigated.
In order to improve the effluent quality, the adopted Stover–Kincannon kinetic model was
coded using MATLAB object-oriented language and the predicted values showed the
applicability of the model. The simulated data were in good agreement with the observed
results, which indicated high correlation of the model to predict effluent COD concentration.
Likewise, the developed MMGM model was used to predict methane production from AD of
industrial wastewater and the results showed the applicability of the developed model to
predict usable methane component of biogas produced during AD of brewery wastewater.
Finally, a combined Pareto multi–objective differential evolution algorithm was successfully
applied to optimize multi-objective anaerobic treatment problem. Its performance was very
encouraging when compared with other multi–objective evolutionary algorithms using
common benchmark tests for the optimization of anaerobic treatment problem. The algorithm
was tested on the multi-objective anaerobic treatment problem using the modified methane
generation and the adopted Stover–Kincannon models. The results of CPMDE algorithm as
compared with some multi-objective evolutionary algorithms reported in the literature in
terms of convergence and diversity showed that CPMDE algorithm was able to solve multi–
objective high dimensional anaerobic treatment problem with few control parameters.
144
The results further showed that the developed CPMDE algorithm was successful in searching
the feasible solution space for good operational conditions of complex biological processes
that involves multi-objective and multiple constraints operation in a closed system. The non-
dominated solutions generated converge to a Pareto optimal front. The algorithm provided a
useful instrument for simultaneous optimization of various operational parameters needed for
successful running of an UASB reactor; improve methane production rate and effluent
quality. Based on the present optimization study, the CPMDE algorithm produced set of
optimal operational conditions to enhance the plant performance without affecting the plant
configuration. Multi-objective optimization using this evolutionary algorithm was shown to
be a good choice for simultaneous optimization of methane production, biomass washout and
effluent substrate concentration during AD of industrial wastewater—using operational
parameters as possible constraints and decision variables.
This work would help industries using AD technology to design, optimize and control their
on-site anaerobic treatment plants for higher efficiency and renewable energy production.
These results will further help reactor operators or environmental engineers to be more aware
of operating parameters for anaerobic reactor, particularly the studied full-scale UASB
reactor. It will help to set-up optimum operational parameters that will enhance the abilities
of the microbial communities in the treatment unit. Furthermore, the results of the benchmark
test using CPMDE algorithm in this study will be a good tool for process control strategies in
AD operation. It is an elegant, convenient and cost effective tool to investigate certain
engineering questions without using physical experimental time and performing expensive
laboratory tests. Thus, the prediction and optimization of methane production and effluent
quality under different operational conditions could improve the microbial community in
order to increase the efficiency of anaerobic bioreactors.
7.1 SIGNIFICANCE AND NOVELTY OF THE RESEARCH FINDINGS
The research findings in this study are significant in that, bacteria and methanogenic
Archaea populations, as well as methane producing gene concentrations were identified
and quantified in the full-scale UASB reactor and then correlated with the reactor‘s
operational parameters. This study provides an insight for the first time into the diversity
145
of the microbial ecology present in the full-scale UASB reactor granules using different
molecular techniques. DNA-based studies (PCR and QPCR) as used in conjunction with
FISH in a complementary manner provided accurate information about active members of
microbial populations or cells present in the reactor.
An AD process model (MMGM) was developed to improve methane production in an
AD system. To the best of our knowledge, MMGM is the first reported developed model
that serves as both predictive and optimizing tools for brewery wastewater treatment plant
in the literature, as well as multi-objective optimization study using a CPMDE algorithm
for simultaneous optimization of methane production, biomass washout and effluent
substrate concentration from anaerobic reactor treating brewery wastewater.
Furthermore, this may be the first study that reported the identification of microbial
communities in the granular sludge taken from the investigated UASB reactor using
different molecular techniques, as well as a model-based multi-objective optimization
study using CPMDE algorithm for brewery wastewater treatment plant in the literature.
This study increases our knowledge of the microbial communities, especially the
methanogens‘ ability to transform intermediate metabolites during the degradation of
organic matter into biogas at the optimum reactor performance. It is hoped that the results
of this study will help in environmental protection and energy generation during AD of
wastewater in South Africa and, thus contribute to a sustainable long-term clean
development mechanism to generate high methane content in a biogas producing UASB
reactor. The captured methane can then be used as fuel, hence mitigating greenhouse gas
emissions in order to obtain a certified emission reduction credit under the Kyoto
Protocol.
7.2 RECOMMENDATIONS
Due to increase in demand for fresh water by both domestic and industrial users, more
work on post treatment of effluent from anaerobic treatment plant using advanced
technologies should be considered to obtain almost zero pollutant discharge hence, reduce
environmental and freshwater contamination. As observed in this study, treated effluent
was still very high in total suspended solids, nitrogen, ammonia and orthophosphate
concentration when compared with the discharge standards, thus, further treatment is
required in this regard (post treatment).
146
Competition between sulphate-reducing bacteria and methanogenic Archaea using
different molecular techniques should be explored, in order to increase methane
production from bioreactors. Further work on using DGGE and high throughput sequence
could also be done to further elucidate the ecology in each compartment of the UASB
reactor.
In order to meet the demand for energy and reduce the consumption of fossil fuel more
work should be carried out on the economical and sustainability of methane for energy
generation. From the clean development mechanisms point of view, the use of
biologically produced methane for energy generation is classified as a 'carbon neutral'
process and the CO2 released during this process is balanced by the CO2 absorbed by
plants during their growth. Therefore, further work should be carried out in this area.
In addition, government should encourage industries with on-site anaerobic treatment
plants that produce biogas to utilize this for energy generation and conversion to
electricity (green electricity) instead of flaring these gases into the atmosphere. This will
help in mitigation of greenhouse gases into the environment by recycling under-utilized
biogas resources.
Calibration and validation of the developed model (MMGM) using laboratory or pilot-
scale processes treating industrial wastewater should be carried out under different
operational conditions. Thereafter, the techno-economic analysis of biogas production in
a full-scale system for energy generation should be carried out, before upgrading to the
full scale system.
147
REFERENCES
Abbasi, T. and Abbasi, S. A. 2012. Formation and impact of granules in fostering clean
energy production and wastewater treatment in upflow anaerobic sludge blanket
(UASB) reactors. Renewable and Sustainable Energy Reviews, 16 (3): 1696-1708.
Abu Qdais, H., Bani Hani, K. and Shatnawi, N. 2010. Modeling and optimization of biogas
production from a waste digester using artificial neural network and genetic
algorithm. Resources, Conservation and Recycling, 54 (6): 359-363.
Acharya, B. K., Mohana, S. and Madamwar, D. 2008. Anaerobic treatment of distillery spent
wash-A study on upflow anaerobic fixed film bioreactor. Bioresource Technology, 99:
4621-4626.
Acharya, B. K., Pathak, H., Mohana, S., Shouche, Y., Singh, V. and Madamwar, D. 2011.
Kinetic modelling and microbial community assessment of anaerobic biphasic fixed
film bioreactor treating distillery spent wash. Water Research, 45 (14): 4248-4259.
Adeniyi, O. D. 2002. Development of a conceptual mathematical model to predict effluent
discharge. Journal of Brewery Effluent Management. FUTA, Minna, Nigeria, 1: 2-3.
Adeyemo, J. and Enitan, A. 2011. Optimization of fermentation processes using evolutionary
algorithms-A review. Scientific Research and Essays, 6: 1464-1472.
Adeyemo, J. and Otieno, F. 2009a. Application of multi-objective differential evolution
algorithm (MDEA) to irrigation planning. In World Environmental and Water
Resources Congress 2009 at Great Rivers, ASCE, 4689-4698.
Adeyemo, J. and Otieno, F. 2009b. Multi-objective differential evolution algorithm for
solving engineering problems. Journal of Applied Sciences, 9 (20): 3652-3661.
Adeyemo, J. and Otieno, F. 2010. Differential evolution algorithm for solving multi-objective
crop planning model. Agricultural Water Management, 97 (6): 848-856.
Adeyemo, J. A., Olofintoye, O. O. and Otieno, F. A. O. 2014. Performance evaluation of
combined pareto multi-objective differential evolution on tuneable multi-objective
test beds. International Jounal of Simulation modeling, DAAAM Int. Vienna.
Ahn, J. H. and Forster, C. F. 2000. Kinetic analyses of the operation of mesophilic and
thermophilic anaerobic filters treating a simulated starch wastewater. Process
Biochemistry, 36: 19–23.
Ahn, Y.-H., Min, K.-S. and Speece, R. E. 2010. Full-scale UASB reactor performance in the
brewery industry. Environmental Technology, 22: 463-476.
Ahring, B. K. 2003. Biomethanation I Springer Berlin Heidelberg, 81.
148
Akarsubasi, A. T., Ince, O., Oz, N. A., Kirdar, B. and Ince, B. K. 2006. Evaluation of
performance, acetoclastic methanogenic activity and archaeal composition of full-
scale UASB reactors treating alcohol distillery wastewaters. Process Biochemistry, 41
(1): 28-35.
Ali, A., Hashmi, H. N., Querashi, I. A. and Saeed, A. 2009. Treatment feasibility of NSSC
pulping effluent using UASB reactor. Electronic Journal of Environmental and Food
Chemistry, 8 (11): 1085-1090.
Ali, M., Siarry, P. and Pant, M. 2012. An efficient Differential Evolution based algorithm for
solving multi-objective optimization problems. European Journal of Operational
Research, 217 (2): 404-416.
Ali Shah, F., Mahmood, Q., Maroof Shah, M., Pervez, A. and Ahmad Asad, S. 2014.
Microbial ecology of anaerobic digesters: The Key Players of Anaerobiosis. The
Scientific World Journal, 21.
Alvarez, J., Ruiz, I., Gomez, M., Presas, J. and Soto, M. 2006. Start-up alternatives and
performance of an UASB pilot plant treating diluted municipal wastewater at low
temperature. Bioresource Technology, 97: 1640–1649.
Amani, T., Nosrati, M., Mousavi, S. M. and Kermanshahi, R. K. 2011. Study of syntrophic
anaerobic digestion of volatile fatty acids using enriched cultures at mesophilic
conditions. International Journal of Environment Science Technology, 8 (1): 83-96.
Amann, R. I., Binder, B. J., Olson, R. J., Chisholm, S. W., Devereux, R. and Stahl, D. A.
1990. Combination of 16S rRNA-targeted oligonucleotide probes with flow
cytometry for analyzing mixed microbial populations. Applied and Environmental
Microbiology, 56: 1919–1925.
Amann, R. I., Ludwig, W. and Schleifer, K. H. 1995. Phylogenetic identification and in situ
detection of individual microbial cells without cultivation. Microbiology Review, 59
(1): 143–169.
Anderson, G. K., Kasapgil, B. and Ince, O. 1996. Microbial kinetics of a membrane
anaerobic reactor system. Environmental Technology, 17: 449-464.
Anderson, I., Ulrich, L. E., Lupa, B., Susanti, D., Porat, I., Hooper, S. D., Lykidis, A.,
Sieprawska-Lupa, M., Dharmarajan, L., Goltsman, E., Lapidus, A., Saunders, E., Han,
C., Land, M., Lucas, S., Mukhopadhyay, B., Whitman, W. B., Woese, C., Bristow, J.
and Kyrpides, N. 2009. Genomic characterization of Methanomicrobiales reveals
three classes of methanogens. PLoS ONE, 4 (6): e5797.
Angira, R. and Babu, B. V. 2006. Optimization of process synthesis and design problems: A
modified differential evolution approach. Chemical Engineering Science, 61: 4707-
4721.
149
APHA–AWWA–WPCF, Standard methods for the examination of water and wastewater.
1998. 20th ed. Washington, DC, USA. American Public Health Association/American
Water Works Association/Water Environment Federation.
Appels, L., Baeyens, J., Degréve, J. and Dewil, R. 2008. Principles and potential of the
anaerobic digestion of waste-activated sludge. Progress in Energy and Combustion
Science, 34: 755–781.
Aquino, S. F. and Stuckey, D. C. 2004. Soluble microbial products formation in anaerobic
chemostats in the presence of toxic compounds. Water Research, 38: 255-266.
Ariesyady, H. D., Ito, T. and Okabe, S. 2007. Functional bacterial and archaeal community
structures of major trophic groups in a full-scale anaerobic sludge digester. Water
Research, 41 (7): 1554-1568.
Arsova, L. 2010. Anaerobic digestion of food waste: Current status, problems and an
alternative product M.S. Degree, Columbia University.
Atashi, H., Ajamein, H. and Ghasemian, S. 2010. Effect of operational and design parameters
on removal efficiency of a pilot-scale UASB reactor in a sugar factory. World Applied
Sciences Journal, 11 (4): 451-456.
Babaee, A. and Shayegan, J. 2011. Anaerobic digestion of vegetable waste. Chemical
Engineering Transactions, 24: 1291-1296.
Babu, B. V., Chakole, P. G. and Mubeen, J. H. S. 2005. Multiobjective differential evolution
(MODE) for optimization of adiabatic styrene reactor. Chemical Engineering Science,
60 (17): 4822–4837.
Babu, B. V. and Chaurasia, A. S. 2003. Optimization of pyrolysis of biomass using
differential evolution approach. In: Proceedings of the Second International
Conference on Computational Intelligence, Robotics, and Autonomous Systems
(CIRAS), Singapore. December 15-18.
Baig, S., Mahmood, Q., Nawab, B., Hussain, A. and Nafees, M. 2010. Assessment of
seasonal variation in surface water quality of Chitral River, North West Frontier
Province (NWFP). Pakistan World Applied Science Journal, 9 (6): 674-680.
Bapteste, É., Brochier, C. and Boucher, Y. 2005. Higher-level classification of the Archaea:
evolution of methanogenesis and methanogens. Archaea, 1 (5): 353-363.
Barampouti, E. M. P., Mai, S. T. and Vlyssides, A. G. 2005. Dynamic modeling of biogas
production in an UASB reactor for potato processing wastewater treatment. Chemical
Engineering Journal, 106 (1): 53-58.
150
Batstone, D., Keller, J., Angelidaki, I., Kalyhuzhnyi, S., Pavlostathis, S., Rozzi, A., Sanders,
W., Siegrist, H. and Vavilin, V. 2002. The IWA Anaerobic Digestion Model No.1
(ADM1). Water and Science Technology, 45 (10): 65-73.
Batstone, D. J., Keller, J., Newell, R. B. and Newland, M. 2000. Modelling anaerobic
degradation of complex wastewater. I: model development. Bioresource Technology,
75: 67-74.
Bello-Osagie, I. O. and Omoruyi, I. M. 2012. Effect of brewery effluent on the
bacteriological and physicochemical properties of Ikpoba River, Edo State, Nigeria.
Journal of Applied Technology in Environmental Sanitation, 2 (4): 197-204.
Bergmann, I. 2012. Characterization of methanogenic Archaea communities in biogas
reactors by quantitative PCR. der Technischen Universität Berlin.
Bhatti, Z. I., Furukawa, K. and Fujita, M. 1993. Current and future trends in water pollution
control in developing countries-A case study of Pakistan. Japanese Journal of Water
Treatment Biology, 29: 51-59.
Bhatti, Z. I., Furukawa, K. and Fujita, M. 1996. Feasibility of methanolic waste treatment in
UASB reactors. Water Research, 30 (11): 2559-2568.
Bhunia, P. and Ghangrekar, M. M. 2008. Analysis, evaluation, and optimization of kinetic
parameters for performance appraisal and design of UASB reactors. Bioresource
Technology, 99 (7): 2132-2140.
Bialek, K., Kim, J., Lee, C., Collins, G., Mahony, T. and O'Flaherty, V. 2011. Quantitative
and qualitative analyses of methanogenic community development in high-rate
anaerobic bioreactors. Water Research, 45(3): 1298-1308.
Bitton, G. 1994. Anaerobic Digestion of Wastewater and Sludge. In: Wastewater
Microbiology. Wiley-Liss, Inc., ISBN 229-245, USA.
Blumensaat, F. and Keller, J. 2002. Modelling of two-stage anaerobic digestion using the
IWA Anaerobic Digestion Model No. 1 (ADM1). Water Research, 39: 171-183.
Bocher, B., Agler, M., Garcia, M., Beers, A. and Angenent, L. 2008. Anaerobic digestion of
secondary residuals from an anaerobic bioreactor at a brewery to enhance bioenergy
generation. Journal of Industrial Microbiology and Biotechnology, 35 (5): 321-329.
Bond, T. and Templeton, M. R. 2011. History and future of domestic biogas plants in the
developing world. Energy for Sustainable Development, 15 (4): 347-354.
Boonlong, K. 2013. Multi-objective optimization of a vehicle vibration model using the
improved compressed-objective genetic algorithm with convergence detection.
Advances in Mechanical Engineering, 2013: 14.
151
Borja, R., Rinc n, B., Raposo, F., Alba, J. and Mart n, A. 2003. Kinetics of mesophilic
anaerobic digestion of the two-phase olive mill solid waste. Biochemical Engineering
Journal, 15 (2): 139-145.
Botheju, D. and Bakke, R. 2011. Oxygen effects in anaerobic digestion-a review. Open Waste
Management Journal, 4.
Briones, A. and Raskin, L. 2003. Diversity and dynamics of microbial communities in
engineered environments and their implications for process stability. Current Opinion
in Biotechnology, 14: 270-276.
Brito, A., Peixoto, J., Oliveira, J., Costa, C., Nogueira, R. and Rodrigues, A. 2007. Brewery
and winery wastewater treatment: some focal points of design and operation. In:
Oreopoulou, V., Russ, W. (Eds.), Utilization of by-products and treatment of waste in
the food industry, ISEKI-Food. Springer.
Britz, T. J., Schalkwyk, C. V. and Roos, P. 2002. Development of a method to enhance
granulation in a laboratory batch system. Water SA, 28: 49- 54.
Brochier-Armanet, C., Boussau, B., Gribaldo, S. and Forterre, P. 2008. Mesophilic
crenarchaeota: proposal for a third archaeal phylum, the Thaumarchaeota. Nature
Reviews Microbiology, 6, 245–252.
Buriánková, I., Brablcová, L., Mach, V., Dvořák, P., Chaudhary, P. P. and Rulík, M. 2013.
Identification of Methanogenic Archaea in the hyporheic sediment of Sitka stream.
PLoS ONE, 8 (11): e80804.
Cakmakci, M. 2007. Adaptive neuro-fuzzy modeling of anaerobic digestion of primary
sedimentation sludge. Bioprocess and Biosystems Engineering, 30 (5): 349-357.
Calli, B., Mertoglu, B., Inanc, B. and Yenigun, O. 2005. Effects of high free ammonia
concentrations on the performances of anaerobic bioreactors. Process Biochemistry,
40: 1285-1292.
Cardinali-Rezende, J., Debarry, R., Colturato, L. D. B., Carneiro, E., Chartone-Souza, E. and
Nascimento, A. A. 2009. Molecular identification and dynamics of microbial
communities in reactor treating organic household waste. Applied Microbiology and
Biotechnology, 84 (4): 777-789.
Casserly, C. and Erijman, L. 2003. Molecular monitoring of microbial diversity in an UASB
reactor. International Biodeterioration and Biodegradation, 52: 7-12.
Castillo, A., Llabres, P. and Alvarez, M. J. 1999. A kinetic study of a combined anaerobic–
aerobic system for treatment of domestic sewage. Water Research, 33: 1742–1747.
152
Castro, H., Ogram, A. and Reddy, K. R. 2004. Phylogenetic characterization of methanogenic
assemblages in eutrophic and oligotrophic Areas of the Florida Everglades. Applied
and Environmental Microbiology, 70 (11): 6559–6568.
Cecchi, F., Mata-Alvarez, J., Pavan, P., Vallini, G. and De Poli, F. 1992. Seasonal effects on
anaerobic digestion of the source sorted organic fraction of municipal solid waste.
Waste Management and Research, 10 (5): 435-443.
Chae, K. J., Jang, A., Yim, S. K. and Kim, I. S. 2007. The effect of digestion temperature and
temperature shock on the biogas yields from the mesophilic anaerobic digestion of
swine manure. Bioresource Technology, 99 (1): 1-6.
Chan, O.-C., Liu, W.-T. and Fang, H. H. P. 2001. Study of microbial community of brewery-
treating granular sludge by denaturing gradient gel electrophoresis of 16S rRNA gene.
Water Science and Technology, 43 (1): 77-82.
Chan, Y. J., M. F., Law, C. L. and Hassell, D. G. 2009. A review on anaerobic–aerobic
treatment of industrial and municipal wastewater. Chemical Engineering Journal, 155
1–18.
Chen, Y., Cheng, J. J. and Creamer, K. S. 2008. Inhibition of anaerobic digestion process: A
review. Bioresource Technology, 99 (10): 4044-4064.
Chen, Y. R. and Hashimoto, A. G. 1978. Kinetics of methane fermentation. Biotechnology
and Bioengineering symposium, 8: 269–282.
Cheng, J. and Liu, B. 2002. Swine wastewater treatment in anaerobic digesters with floating
medium. Transactions of the American Society of Agricultural Engineers (ASAE), 45
(3): 799–805.
Cheng, L., Dai, L., Li, X., Zhang, H. and Lu, Y. 2011. Isolation and characterization of
Methanothermobacter crinale sp. nov, a novel hydrogenotrophic methanogen from
Shengli Oil fields. Applied and Environmental Microbiology, 77: 5212-5219.
Chong, S., Sena, T. K., Kayaalp, A. and Ang, H. M. 2012. The performance enhancements of
upflow anaerobic sludge blanket (UASB) reactors for domestic sludge treatment–A
State-of-the-art review. Water Research, 46 (11): 3434–3470.
Chouari, R., Paslier, D. L., Daegelen, P., Dauga, C., Weissenbach, J. and Sghir, A. 2010.
Molecular analyses of the microbial community composition of an anoxic basin of a
municipal wastewater treatment plant reveal a novel lineage of Proteobacteria.
Microbial Ecology, 60 (2): 272-281.
Chulhwan, P., Chunyeon, L., Sangyong, K., Yu, C. and Howard, C. H. 2005. Upgrading of
anaerobic digestion by incorporating two different hydrolysis processes. Journal of
Bioscience and Bioengineering, 100: 164–167.
153
Clara, M., Kreuzinger, N., Strenn, B., Gans, O. and Kroiss, H. 2005. The solids retention
time-a suitable design parameter to evaluate the capacity of wastewater treatment
plants to remove micro pollutants. Water Research, 39: 97–106.
Cloete, R. 2008. Biogas train on track for completion. Engineering news, renewable energy
(http://www.talbot.co.za/wp-content/uploads/2012/12/Biogas_train.pdf). July 25–31
2008).
Coelho, N., Rodrigues, A., Arroja, L. and Capela, I. 2007. Effect of non-feeding period
length on the intermittent operation of UASB reactors treating dairy effluents.
Biotechnology Bioengineering, 96 (2): 244-249.
Colussi, I., Cortesi, A., Gallo, V., Fernandez, A. S. R. and Vitanza, R. 2012. Modelling of an
anaerobic process producing biogas from winery wastes. Chemical Engineering
Transactions, 27: 301-306.
Contois, D. E. 1959. Kinetics of microbial growth: Relationship between population density
and specific growth rate of continuous culture. Journal of General Microbiology, 21
(1): 40-50.
Crocetti, G., Murto, M. and Björnsson, L. O. 2006. An update and optimisation of
oligonucleotide probes targeting methanogenic Archaea for use in fluorescence in-situ
hybridisation (FISH). Journal of Microbiology Methods, 65: 194-201.
Cronin, C. and Lo, K. V. 1998. Anaerobic treatment of brewery wastewater using UASB
reactors seeded with activated sludge. Bioresource Technology, 64: 33-38.
Daims, H., Bruhl, A., Amann, R., Schleifer, K.-H. and Wagner, M. 1999. The domain-
specific probe EUB338 is insufficient for the detection of all Bacteria: Development
and evaluation of a more comprehensive probe set. Systematic and Applied
Microbiology, 22: 434-444.
Danazumi, S. and Hassan, M. 2010. Industrial pollution and implication on source of water
supply in Kano, Nigeria. International Journal of Engineering & Technology, 10 (1):
101-109.
Dar, S. A., Kleerebezem, R., Stams, A. J. M., Kuenen, J. G. and Muyzer, G. 2008.
Competition and coexistence of sulfate-reducing bacteria, acetogens and methanogens
in a lab-scale anaerobic bioreactor as affected by changing substrate to sulfate ratio.
Applied Microbiology and Biotechnology, 78 (6): 1045–1055.
Das, S., Abraham, A. and Konar, A. 2008. Particle swarm optimization and differential
evolution algorithms: Technical analysis, applications and hybridization perspectives.
Studies in Computational Intelligence (SCI), 116: 1–38. www.springerlink.com.
154
Deb, K. 2011. Multi-objective optimization using evolutionary algorithms: An introduction.
multi-objective evolutionary optimisation for product design and manufacturing. 1-24.
Deb, K. 2001. Multi-objective optimization using evolutionary algorithms. First edition ed.
Chichester, UK: John Wiley & Sons LTD.
Deb, K., Agrawal, S., Pratap, A. and Meyarivan, T. 2000. A fast elitist non-dominated sorting
genetic algorithm for multi-objective optimization: NSGA-II. Lecture notes in
computer science, 1917: 849-858.
Debik, E. and Coskun, T. 2009. Use of the static granular bed reactor (SGBR) with anaerobic
sludge to treat poultry slaughterhouse wastewater and kinetic modelling. Bioresource
Technology, 100: 2777–2782.
Delbès, C., Moletta, R. and Godon, J.-J. 2001. Bacterial and archaeal 16S rDNA and 16S
rRNA dynamics during an acetate crisis in an anaerobic digestor ecosystem. FEMS
Microbiology Ecology, 35 (1): 19-26.
Demirel, B. and Scherer, P. 2008a. Production of methane from sugar beet silage without
manure addition by a single-stage anaerobic digestion process. Biomass and
Bioenergy, 32: 203-209.
Demirel, B. and Scherer, P. 2008b. The roles of acetotrophic and hydrogenotrophic
methanogens during anaerobic conversion of biomass to methane: a review. Reviews
in Environmental Science and Biotechnology, 7: 173–190.
Demirel, B., Scherer, P., Yenigun, O. and Onay, T. T. 2010. Production of methane and
hydrogen from biomass through conventional and high-rate anaerobic digestion
processes. Critical Reviews in Environmental Science and Technology, 40: 116–146.
Deng, W., Yang, X., Zou, L., Wang, M., Liu, Y. and Li, Y. 2013. An improved self-adaptive
differential evolution algorithm and its application. Chemometrics and Intelligent
Laboratory Systems, 128: 66-76.
Department of Public Works Republic of South Africa. 2012. Small domestic wastewater
treatment plant guideline, 18.
Derbal, K., Bencheikh-Iehocine, M., Cecchi, F., Meniai, A. H. and Pavan, P. 2009.
Application of the IWA ADM1 model to simulate anaerobic co-digestion of organic
waste with waste activated sludge in mesophilic condition. Bioresource Technology,
100: 1539-1543.
Deshmukh, M. K. and Moorthy, C. B. 2010. Application of genetic algorithm to neural
network model for estimation of wind power potential. Journal of Engineering,
Science and Management Education, 2: 42-48.
155
Diamantis, V. I. and Alexandros, A. 2007. Comparison of single- and two stage UASB
reactors used for anaerobic treatment of synthetic fruit wastewater. Enzyme and
Microbial Technology, 42: 6–10.
Diaz, E. E., Stams, A. J. M., Amils, R. and Sanz, J. L. 2006. Phenotypic properties and
microbial diversity of methanogenic granules from a full scale upflow anaerobic
sludge bed reactor treating brewery wastewater. Applied and Environmental
Microbiology, 72 (7): 4942-4949.
Dojka, M. A., Hugenholtz, P., Haack, S. K. and Pace, N. R. 1998. Microbial diversity in a
hydrocarbon- and chlorinated-solvent-contaminated aquifer undergoing intrinsic
bioremediation. Applied and Environmental Microbiology, 64 (10): 3869-3877.
Driessen, W. and Vereijken, T. 2003. Recent developments in biological treatment of
brewery effluent. The Institute and Guild of Brewing Convention, Livingstone,
Zambia, March 2-7.
Droste, R. L. 1997. Theory and practice of water and wastewater treatment Talanta, (Chapter
18): 622-635.
Dunaj, S. J., Vallino, J. J., Hines, M. E., Gay, M., Kobyljanec, C. and Rooney-Varga, J. N.
2012. Relationships between soil organic matter, nutrients, bacterial community
structure, and the performance of microbial fuel cells. Environmental Science and
Technology, 46 (3): 1914-1922.
Dutschke, M., Kapp, G., Lehmann, A. and Schafer, V. 2006. Risks and chances of combined
forestry and biomass projects under the Clean Development Mechanism: UNEP Riso
Centre.
Elnekave, M., Celik, S. O., Tatlier, M. and Tufekci, N. 2012. Artificial neural network
predictions of up-flow anaerobic sludge blanket (UASB) reactor performance in the
treatment of citrus juice wastewater. Polish Journal of Environmental Studies, 21 (1):
49-56.
Enitan, A. M. and Adeyemo, J. 2011. Food processing optimization using evolutionary
algorithms. African Journal of Biotechnology, 10 (72): 16120-16127.
Enitan, A. M., Kumari, S., Swalaha, F. M., Adeyemo, J., Ramdhani, N. and Bux, F. 2014a.
Kinetic modelling and characterization of microbial community present in a full-scale
UASB reactor treating brewery effluent. Microbial Ecology, 67: 358–368.
Enitan, A. M., Swalaha, F. M., Adeyemo, J. and Bux, F. 2014b. Assessment of brewery
effluent composition from a beer producing industry In KwaZulu–Natal, South
Africa. Fresenius Environmental Bulletin, 23 (3): 693-701. 1).
156
Ericson, E., Thorin, E. and Yan, J. 2010. Exploring the possibility of using a simple neural
network for the prediction of biogas production of a solid waste digester, poster
presentation. Paper presented at October 31st – November 4th, 2010 Guadalajara,
Mexico.
Espinoza-Escalante, F. M., Pelayo-Ort z, C., Navarro-Corona, J., Gonzalez-Garc a, Y.,
Bories, A. and Gutierrez-Pulido, H. 2009. Anaerobic digestion of the vinasses from
the fermentation of agave tequilana Weber to tequila: The effect of pH, temperature
and hydraulic retention time on the production of hydrogen and methane. Biomass
and Bioenergy, 33: 14-20.
Estes, L., Bradley, B., Beukes, H., Hole, D., Lau, M., Oppenheimer, M., Schulze, R.,
Tadross, M. and Turner, W. 2013. Comparing mechanistic and empirical model
projections of crop suitability and productivity: implications for ecological
forecasting. Global Ecology and Biogeography, 22(8): 1007-1018.
Faisal, M. and Unno, H. 2001. Kinetic analysis of palm oil mill wastewater treatment by a
modified anaerobic baffled reactor. Biochemical Engineering Journal, 9: 25-31.
Fang, H. H. P., Chui, H.-K. and Li, Y.-Y. 1995a. Anaerobic degradation of butyrate in a
UASB reactor. Bioresource Technology, 51 (1): 75-81.
Fang, H. H. P., Chui, H. K. and Li, Y. Y. 1994. Microbial structure and activity of UASB
granules treating different wastewaters. Proceeding of 7th International Symposium
on Anaerobic Digestion, 23-27 January, Cape Town, South Africa: 80-89.
Fang, H. H. P., Chui, H. K. and Li, Y. Y. 1995b. Microstructural analysis of UASB granules
treating brewery wastewater. Water Science and Technology, 31: 129-135.
Fatoki, O. S. and Mathabatha, S. 2001. An assessment of heavy metal pollution in the East
London and Port Elizabeth harbors. Water SA, 27: 233–240.
Fdez-Güelfo, L. A., Álvarez-Gallego, C., Sales, D. and Romero García, L. I. 2012. Dry-
thermophilic anaerobic digestion of organic fraction of municipal solid waste:
Methane production modeling. Waste Management, 32 (3): 382-388.
Fermoso, F., Collins, G., Bartacek, J., O‘Flaherty, V. and Lens, P. 2008. Role of nickel in
high rate methanol degradation in anaerobic granular sludge bioreactors.
Biodegradation, 19 (5): 725-737.
Ferry, J. 1993. Methanogenesis: ecology, physiology, biochemistry and genetics. New York:
Chapman & Hill.
Ficker, M., Krastel, K., Orlicky, S. and Edwards, E. 1999. Molecular characterization of a
toluene-degrading methanogenic consortium. Applied and Environmental
Microbiology, 65 (12): 5576-5585.
157
Fillaudeau, L., Blanpain-Avet, P. and Daufin, G. 2006. Water, wastewater and waste
management in brewing industries. Journal of Cleaner Production, 14: 463-471.
Fister, I., Yang, X.-S., Brest, J. and Fister Jr, I. 2013. Modified firefly algorithm using
quaternion representation. Expert Systems with Applications, 40 (18): 7220-7230.
Gallert, C. and Winter, J. 2008. Propionic acid accumulation and degradation during restart of
a full scale anaerobic biowaste digester. Bioresource Technology, 99: 170–178.
Garcia, M. L. and Angenent, L. T. 2009. Interaction between temperature and ammonia in
mesophilic digesters for animal waste treatment. Water Research, 43: 2373-2382.
Gerardi, M. H. 2003. The Microbiology of Anaerobic Digesters. Wiley-Interscience, New
Jersey, 51-57.
Ghaly, A. E., Sadaka, S. S. and Hazza'a, A. 2000. Kinetics of an intermittent-flow,
continuous-mix anaerobic reactor. Energy Sources, 22 (6): 525-542.
Ghosh, D. 2013. Improving the Microbial Production of Biofuels through Metabolic
Engineering de Philosophiae Doctor, Université de Montréal.
Giovannoni, S. J. 1991. The polymerase chain reaction. In:Stackebrandt E, Good fellow M
(eds) Nucleic Acid Techniques in Bacterial Systematics. John Wiley and Sons,
Chischester, 177-203.
Gomec, C. Y., Letsiou, I., Ozturk, I., Eroglu, V. and Wilderer, P. A. 2008. Identification of
Archaeal population in the granular sludge of an UASB reactor treating sewage at low
temperatures. Journal of Environmental Science and Health Part A, 43: 1504–1510.
Gomes, J., Singhal, A., Praveen, V. and Ramachandran, K. 1998. Axial dispersion model for
upflow anaerobic sludge blanket reactors. Biotechnology Progress, 14: 645-648.
Goñi, S. M., Oddone, S., Segura, J. A., Mascheroni, R. H. and Salvadori, V. O. 2008.
Prediction of foods freezing and thawing times: Artificial neural networks and genetic
algorithm approach. Journal of Food Engineering, 84 (1): 164-178.
Gonzalez, J. M., Ortiz-Martinez, A., Gonzalez-delValle, M. A., Laiz, L. and Saiz-Jimenez, C.
2003. An efficient strategy for screening large cloned libraries of amplified 16S
rDNA sequences from complex environmental communities. Journal of
Microbiological Methods, 55: 459-463.
Govahi, S., Karimi-Jashni, A. and Derakhshan, M. 2012. Treatability of landfill leachate by
combined upflow anaerobic sludge blanket reactor and aerated lagoon. International
Journal of Environmental Science and Technology, 9 (1): 145-151.
158
Grothenhuis, J. T. C., Smith, M., Plugge, C. M., Yuansheng, X., Lammeren, A. A. M., Stams,
A. J. M. and Zehnder, A. J. B. 1991. Bacteriological composition and structure of
granular sludge adapted to different substrates. Applied and Environmental
Microbiology, 57: 1942-1949.
Gueguim Kana, E. B., Oloke, J. K., Lateef, A. and Adesiyan, M. O. 2012. Modeling and
optimization of biogas production on saw dust and other co-substrates using artificial
neural network and genetic algorithm. Renewable Energy, 46: 276-281.
Gunaseelan, V. 2007. Regression models of ultimate methane yields of fruits and vegetable
solid wastes, sorghum and Napier grass on chemical composition. Bioresource
Technology, 98: 1270–1277.
Gyalpo, T., Vögeli, Y. and Zurbrügg, C. 2008. Quantification of methane emissions from
uncontrolled dumping of solid waste and from different sanitation systems in
developing countries. Institute of Biogeochemistry and Pollutant Dynamics
Department Environmental Sciences, ETH Zürich December 2008.
www.ibp.ethz.ch//HS08_Gyalpo_rev_termpaper_ms.pdf,(Accessed date: 05/11/2012).
Habeeb, S. A., Latiff, A. A. A., Daud, Z. and Ahmad, Z. 2011. The start-up of hybrid,
anaerobic up-flow sludge blanket (HUASB) under a range of mesophilic and
thermophilic temperatures. Environment Asia, 4 (2): 63-68.
Hampannavar, U. S. and Shivayogimath, C. B. 2010. Anaerobic treatment of sugar industry
wastewater by upflow anaerobic sludge blanket reactor at ambient temperature.
International Journal of Environmental Sciences, 1(4): 631-639.
He, J., Ritalahti, K. M., Aiello, M. R. and Loffler, F. E. 2003. Complete detoxification of
vinyl chloride by an anaerobic enrichment culture and identification of the reductively
dechlorinating population as a Dehalococcoides species. Applied and Environmental
Microbiology, 69 (2): 996–1003.
Heffernan, B., Blanc, J. and Spanjers, H. 2012. Evaluation of greenhouse gas emissions from
municipal UASB sewage treatment plants. Journal of Integrative Environmental
Sciences, 9 (Supplement 1): 127-137.
Herumurti, W., Isa, M. H. and Kutty, S. R. M. 2008. Treatment of pharmaceutical wastewater
using mesophilic UASB and hybrid UASB reactors. 4th International Conference on
Sustainable Water Environment: Innovative Technologies and Energy Efficient
Solutions. Singapore, 17 – 19 November 2008.
Hoffmann, J. R. H. 1985. High rate anaerobic treatment of brewery wastes. Paper presented
at SA Institution of Chemical Engineers (Natal Branch) Meeting, 28 February.
Hofman-Bang, J., Zheng, D., Westermann, P., Ahring, A. and Raskin, K. 2003. Molecular
ecology of anaerobic reactor systems. Springer-Verlag GmbH, Heidelberg.
159
Holland, J. H. 1975. Adaptation in Natural and Artificial Systems. University of Michigan
Press, Ann Arbe, MI.
Holubar, P., Zani, L., Hagar, M., Froschl, W., Radak, Z. and Braun, R. 2002. Advanced
controlling of anaerobic digestion by means of hierarchical neural networks. Water
Research, 36 (10): 2582-2588.
Huang, V. L., Suganthan, P. N., Qin, A. K. and Baskar, S. 2005. Multi-objective differential
evolution with external archive and harmonic distance-based diversity measure.
School of Electrical and Electronic Engineering Nanyang, Technological University
Technical Report: 2005.
Huang, Z. and Chen, Y. 2013. An Improved Differential Evolution Algorithm Based on
Adaptive Parameter. Journal of Control Science and Engineering, Volume 2013,
Article ID 462706, 5 pages http://dx.doi.org/10.1155/2013/462706. Hindawi
Publishing Corporation.
Hulshoff-Pol, H. 1995. Waste characteristics and factors affecting reactor performance.
Lecture notes by Hulshoff Pol in International Course on Anaerobic Wastewater
Treatment, Wageningen Agriculture University, The Delft, Netherlands.
Hulshoff-Pol , L. W. H., Lopes, S. I. D. C., Lettinga, G. and Lens, P. N. L. 2004. Anaerobic
sludge granulation. Water Research, 38: 1376-1389.
Ikhu-Omoregbe, D., Kuipa, P. K. and Hove, M. 2005. An assessment of the quality of liquid
effluents from opaque beer-brewing plants in Bulawayo, Zimbabwe. Water SA, 31
(1): 141-150.
Inanc, B., Calh, B. and Saatci, A. 2000. Characterisation and anaerobic treatment of the
sanitary landfill leachate in Istanbul. Water Science and Technology, 41 (3): 223–230.
Inanc, B., Matsui, S. and Ide, S. 1999. Propionic acid accumulation in anaerobic digestion of
carbohydrates: an investigation on the role of hydrogen gas. Water Science and
Technology, 40 (1): 93–100.
Ince, B. K., Ayman-Oz, N., Türker, G., Celikkol, S. and Ince, O. 2010. Microbial ecology of
anaerobic reactors for treatment of alcohol industry wastewaters: a review. Current
Research, Technology and Education Topics in Applied Microbiology and Microbial
Biotechnology, 988-999.
Inyang, U. E., Bassey, E. N. and Inyang, J. D. 2012. Characterization of brewery effluent
fluid. Journal of Engineering and Applied Sciences, 4: 67-77.
International Energy Agency (IEA). 2001. Biogas and More! Systems and Markets Overview
of Anaerobic digestion.
160
Ipeaiyeda, A. R. and Onianwa, P. C. 2012. Impact of brewery effluent on water quality of the
Olosun River in Ibadan, Nigeria. Chemistry and Ecology, 25 (3): 189-204.
Iqbal, J. and Guria, C. 2009. Optimization of an operating domestic wastewater treatment
plant using elitist non-dominated sorting genetic algorithm. Chemical Engineering
Research and Design, 87 (11): 1481-1496.
Isherwood, H. A. 1991. The treatment of brewery wastewater using anaerobic digestion-
Some operational experiences. Paper presented at the 3rd Central and Southern Africa
Institute of Brewing Convention. Victoria Falls, Zimbabwe. 3-7 March, 243-249.
Islam, M., Khan, H., Das, A., Akhtar, M., Oki, Y. and Adochi, T. 2006. Impacts of industrial
effluents on plant growth and soil properties. Soil and Environment, 25 (2): 113-118.
Iversen, C., Mullane, N., McCardell, B., Tall, B. D., Lehner, A., Fanning, S. A., Stephan, R.
and Joosten, H. 2008. Cronobacter gen. nov., a new genus to accommodate the
biogroups of Enterobacter sakazakii, and proposal of Cronobacter sakazakii gen.
nov., comb. nov., Cronobacter malonaticus sp. nov., Cronobacter turicensis sp. nov.,
Cronobacter muytjensii sp. nov., Cronobacter dublinensis sp. nov., Cronobacter
genomospecies 1, and of three subspecies, Cronobacter dublinensis subsp.
dublinensis subsp. nov., Cronobacter dublinensis subsp. lausannensis subsp. nov. and
Cronobacter dublinensis subsp. lactaridi subsp. nov. International Journal of
Systematic and Evolutionary Microbiology, 58: 1442–1447.
Jang, H. M., Kim, J. H., Ha, J. H. and Park, J. M. 2014. Bacterial and methanogenic archaeal
communities during the single-stage anaerobic digestion of high-strength food
wastewater. Bioresource Technology, 165 (2014): 174-182.
Jeong, H.-S., Suh, C.-W., Lim, J.-L. and Shin, H.-S. 2005. Analysis and application of
ADM1 for anaerobic methane production. Bioprocess and Biosystem Engineering, 27:
81-89.
Jones, C. L. W., Britz, P., Davies, M. T. T., Scheepers, R., Cilliers, A., Crous, L. and
Laubscher, R. 2011. The wealth in brewery effluent – Water and nutrient recovery
using alternative technologies In: Proceedings of Fifteenth International Water
Technology Conference, IWTC-15 2011. Alexandria, Egypt.
Joulian, C., Patel, B. K., Ollivier, B., Garcia, J. L. and Roger, P. A. 2000. Methanobacterium
oryzae sp. nov., a novel methanogenic rod isolated from a Philippines rice field.
International Journal of Systematic and Evolutionary Microbiology, 50 (2): 525-528.
Jupraputtasri, W., Boonapatcharoen, N., Cheevadhanarak, S., Chaiprasert, P., Tanticharoen,
M. and Techkarnjanaruk, S. 2005. Use of an alternative Archaea-specific probe for
methanogen detection. Journal of Microbiological Methods, 61 (1): 95-104.
161
Kachitvichyanukul, V. 2012. Comparison of Three Evolutionary Algorithms: GA, PSO, and
DE. Industrial Engineering and Management Systems, 11(3): 215-223.
Kampmann, K., Ratering, S., Baumann, R., Schmidt, M., Zerr, W. and Schnell, S. 2012.
Hydrogenotrophic methanogens dominate in biogas reactors fed with defined
substrates. Systematic and Applied Microbiology, 35 (6): 404-413.
Kanat, G. and Saral, A. 2009. Estimation of biogas production rate in a thermophilic UASB
reactor using artificial neural networks. Environmental Modeling and Assessment, 14
(5): 607-614.
Kang, J.-H., Kim, D. and Lee, T.-J. 2012. Hydrogen production and microbial diversity in
sewage sludge fermentation preceded by heat and alkaline treatment. Bioresource
Technology, 109: 239-243.
Kanu, I. and Achi, O. 2011. Industrial effluents and their impact on water quality of receiving
rivers in Nigeria. Journal of Applied Technology in Environmental Sanitation, 1 (1):
75-86.
Kaparaju, P., Serrano, M. and Angelidaki, I. 2010. Optimization of biogas production from
wheat straw stillage in UASB reactor. Applied Energy, 87 (12): 3779-3783.
Kapdan, I. K. and Erten, B. 2007. Anaerobic treatment of saline wastewater by
Halanaerobium lacurosei. Process Biochemistry, 42: 449–453.
Karaboga, D. 2004. A simple and global optimization algorithm for engineering problems:
differential evolution algorithm. Turkish Journal of Electrical Engineering, 12 (1):
53-60.
Karagiannidis, A. and Perkoulidis, G. 2009. A multi-criteria ranking of different technologies
for the anaerobic digestion for energy recovery of the organic fraction of municipal
solid wastes. Bioresource Technology, 100: 2355–2360.
Karakashev, D., Batstone, D. J., Trably, E. and Angelidaki, I. 2006. Acetate oxidation is the
dominant pathway from acetate in the absence of Methanosaetaceae. Applied and
Environmental Microbiology, 72(7): 5138–5141.
Karnholz, A., Kusel, K., Gossner, A., Schramm, A. and Drake, H. L. 2002. Tolerance and
metabolic response of acetogenic bacteria toward oxygen. Applied and Environmental
Microbiology, 68: 1005-1009.
Keyser, M. 2006. PCR detection, denaturing gradient gel electrophoresis (DGGE)
fingerprinting and identification of the microbial consortium in different types of
UASB granules. Doctor of Philosophy in Food Science, University of Stellenbosch.
162
Keyser, M., Witthuhn, R. C., Lamprecht, C., Coetzee, M. P. A. and Britz, T. J. 2006. PCR-
based DGGE fingerprinting and identification of methanogens detected in three
different types of UASB granules. Systematic and Applied Microbiology, 29 (1): 77-
84.
Keyser, M., Witthuhn, R. C., Ronquest, L. C. and Britz, T. J. 2003. Treatment of winery
effluent with upflow anaerobic sludge blanket (UASB)—granular sludges enriched
with Enterobacter sakazakii. Biotechnology Letter, 25 (22): 1893–1898.
Khalid, A., Arshad, M., Anjum, M., Mahmood, T. and Dawson, L. 2011. The anaerobic
digestion of solid organic waste. Waste Management, 31 (8): 1737-1744.
Khataee, A. R. and Kasiri, M. B. 2011. Modeling of biological water and wastewater
treatment processes using artificial neural networks. Clean – Soil, Air, Water, 39 (8):
742–749.
Khemkhao, M., Nuntakumjorn, B., Techkarnjanaruk, S. and Phalakornkule, C. 2012. UASB
performance and microbial adaptation during a transition from mesophilic to
thermophilic treatment of palm oil mill effluent. Journal of Environmental
Management, 103 (0): 74-82.
Khosla, D. K., Gupta, S. K. and Saraf, D. N. 2007. Multi objective optimization of fuel oil
blending using the jumping gene adaptation of genetic algorithm. Fuel Process
Technology, 88: 51–63.
Kilani, J. S. 1993. A compatibility study of the effects of dairy and brewery effluents on the
treatability of domestic sewage. Water SA, 19 (3): 247-252.
Kim, J. K., Oh, B. R., Chun, Y. N. and Kim, S. W. 2006. Effects of temperature and
hydraulic retention time on anaerobic digestion of food waste. Journal of Bioscience
and Bioengineering, 102: 328-332.
Kincannon, D. F. and Stover, E. L. 1982. Design methodology for fixed film reaction-RBCs
and biological towers. New York: Pergamon Press.
Kirin Holdings. 2012. Kirin Institute of Food and Lifestyle Report, Global Beer Production
by Country in 2011 (http://www.kirinholdings.co.jp/english/news/2012/0808_01.html
(Accessed: March 03, 2013).
Klocke, M., Mahnerta, P., Mundta, K., Souidi, K. and Linke, B. 2007. Microbial community
analysis of a biogas-producing completely stirred tank reactor fed continuously with
fodder beet silage as mono-substrate. Systematic and Applied Microbiology, 30: 139–
151.
163
Kovacik, W. P. J., Scholten, J. C. M., Culley, D., Hickey, R., Zhang, W. and Brockman, F. J.
2010. Microbial dynamics in upflow anaerobic sludge blanket (UASB) bioreactor
granules in response to short-term changes in substrate feed. Microbiology, 156:
2418–2427.
Kovoor, P. P., Idris, M. R., Hassan, M. H. and Yahya, T. F. T. 2012. A study conducted on
the impact of effluent waste from machining process on the environment by water
analysis. International Journal of Energy and Environmental Engineering, 3:21.
Krakat, N., Schmidt, S. and Scherer, P. 2010. Mesophilic fermentation of renewable biomass:
Does hydraulic retention time regulate methanogen diversity? Applied and
Environmental Microbiology, 76 (18): 6322-6326.
Krakat, N., Schmidt, S. and Scherer, P. 2011. Potential impact of process parameters upon the
bacterial diversity in the mesophilic anaerobic digestion of beet silage. Bioresource
Technology, 102 (10): 5692-5701.
Krishna, R. H. 2013. Role of factors influencing on anaerobic process for production of bio
hydrogen: Future fuel. International Journal of Advanced Chemistry, 1 (2): 31-38.
Krober, M., Bekel, T., Diaz, N., Goesmann, A., Jaenicke, S., Krause, L., Miller, D., Runte,
K., Viohover, P., Puhler, A. and Schluter, A. 2009. Phylogenetic characterization of a
biogas plant microbial community integrating clone library 16S-rDNA sequences and
metagenome sequence data obtained by 454-pyrosequencing. Journal of
Biotechnology, 142: 38 - 49.
Krzysztof, Z. and Frac, M. 2012. Methane fermentation process as anaerobic digestion of
biomass: Transformations, stages and microorganisms. African Journal of
Biotechnology, 11: 4127.
Kubota, K., Hayashi, M., Matsunaga, K., Iguchi, A., Ohashi, A., Li, Y.-Y., Yamaguchi, T.
and Harada, H. 2014. Microbial community composition of a down-flow hanging
sponge (DHS) reactor combined with an up-flow anaerobic sludge blanket (UASB)
reactor for the treatment of municipal sewage. Bioresource Technology, 151 (0): 144-
150.
Kucerova, E., Clifton, S. W., Xia, X.-Q., Long, F., Porwollik, S., Fulton, L., Fronick, C.,
Minx, P., Kyung, K., Warren, W., Fulton, R., Feng, D., Wollam, A., Shah, N.,
Bhonagiri, V., Nash, W. E., Hallsworth-Pepin, K., Wilson, R. K., McClelland, M. and
Forsythe, S. J. 2010. Genome sequence of Cronobacter sakazakii BAA-894 and
comparative genomic hybridization analysis with other Cronobacter species. PLoS
ONE, 5 (3): 1-10.
Kusiak, A., Zheng, H. Y. and Song, Z. 2009. Wind farm power prediction: A data-mining
approach. Wind Energy, 12 (3): 275-293.
164
Lakshmi, G., Rao, C. S., Rao, R. S., Hobbs, P. and Prakasham, R. 2009. Enhanced production
of xylanase by a newly isolated Aspergillus terreus under solid state fermentation
using palm industrial waste: a statistical optimization. Biochemical Engineering
Journal, 48: 51-57.
Langenhoff, A. A. M. and Stuckey, D. C. 2000. Treatment of a dilute wastewater using an
anaerobic baffled reactor: Effect of low temperature. Water Research, 34 (15): 3867-
3875.
Latif, M. A., Ghufran, R., Wahid, Z. A. and Ahmad, A. 2011. Integrated application of
upflow anaerobic sludge blanket reactor for the treatment of wastewaters. Water
Research, 45: 4683-4699.
Leclerc, M., Delgènes, J. and Godon, J. 2004. Diversity of the archaeal community in 44
anaerobic digesters as determined by single strand conformation polymorphism
analysis and 16S rDNA sequencing. Environmental Microbiology, 6 (8): 809-819.
Lee, S.-H., Kang, H.-J., Lee, Y. H., Lee, T. J., Han, K., Choi, Y. and Park, H.-D. 2012.
Monitoring bacterial community structure and variability in time scale in full-scale
anaerobic digesters. Journal of Environmental Monitoring, 14 (7): 1893-1905.
Lee, S. H., Jung, J. Y. and Jeon, C. O. 2014. Effects of temperature on microbial succession
and metabolite change during saeu-jeot fermentation. Food Microbiology, 38: 16-25.
Leitao, R. C. 2004. Robustness of UASB reactors treating sewage under tropical conditions.
Ph.D. Thesis. Wageningen University.
Leitao, R. C., Haandel, A. C. v., Zeeman, G. and Lettinga, G. 2006. The effects of operational
and environmental variations on anaerobic wastewater treatment systems: A review.
Bioresource Technology, 97: 1105–1118.
Lettinga, G. 1995. Anaerobic digestion and wastewater treatment systems. Antonie Van
Leeuwenhoek, 67 (1): 3-28.
Lettinga, G. and Hulshoff-Pol, L. 1991. UASB-Process design for various types of
wastewaters. Water Science and Technology, 24 (8): 87-107.
Lettinga, G., van Velsen, A. F. M., Hobma, S. W., de Zeeuw, W. and Klapwijk, A. 1980. Use
of the upflow sludge blanket (USB) reactor concept for biological wastewater
treatment, especially for anaerobic treatment. Biotechnology and Bioengineering, 22
(4): 699-734.
Levstek, T. and Lakota, M. 2012. The use of artificial neural networks for compounds
prediction in biogas from anaerobic digestion – A review. Agricultura, 7: 15-22.
165
Li, J., Yu, L., Yu, D., Wang, D., Zhang, P. and Ji, Z. 2014. Performance and granulation in an
upflow anaerobic sludge blanket (UASB) reactor treating saline sulfate wastewater.
Biodegradation, 25 (1): 127-136.
Li, Y., Park, S. Y. and Zhu, J. 2011. Solid-state anaerobic digestion for methane production
from organic waste. Renewable and Sustainable Energy Reviews, 15: 821–826.
Liu, P.-K. and Wang, F.-S. 2008. Inference of biochemical network models in S-system using
multi-objective optimization approach. Bioinformatics, 24 (8): 1085-1092.
Liu Pang-Kai and Wang, F.-S. 2010. Hybrid differential evolution including geometric mean
mutation for optimization of biochemical systems. Journal of the Taiwan Institute of
Chemical Engineers, 41: 65–72.
Liu, W.-T., Chan, O.-C. and Fang, H. H. P. 2002a. Characterization of microbial community
in granular sludge treating brewery wastewater. Water Research, 36 (7): 1767-1775.
Liu, W. T., Chan, O. C. and Fang, H. H. P. 2002b. Microbial community dynamics during
start-up of acidogenic reactors. Water Research, 36: 3203-3210.
Liu, Y. and Tay, H. J. 2002. The essential role of hydrodynamic shear force in the formation
of biofilm and granular sludge. Water Research, 36: 1653–1665.
Liu, Y., Xu, H.-L., Yang, S.-F. and Tay, J.-H. 2003. Mechanisms and models for anaerobic
granulation in upflow anaerobic sludge blanket reactor. Water Research, 37 (3): 661-
673.
Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. 1951. Protein measurement
with the folin phenol reagent. Journal of Biological Chemistry, 193 (1): 265-275.
Lübken, M., Wichern, M., Schlattmann, M., Gronauer, A. and Horn, H. 2007. Modelling the
energy balance of an anaerobic digester fed with cattle manure and renewable energy
crops. Water Research, 41: 4085-4096.
Luton, P. E., Wayne, J. M., Sharp, R. J. and Riley, P. W. 2002. The mcrA gene as an
alternative to 16S rRNA in the phylogenetic analysis of methanogen populations in
landfill. Microbiology, 148: 3521–3530.
Lyberatos, G. and Skiadas, I. V. 1999. Modelling of anaerobic digestion -A review. Global
Nest: The International Journal, 1 (2): 63-76.
Lykidis, A., Chen, C.-L., Tringe, S. G., McHardy, A. C., Copeland, A., Kyrpides, N. C.,
Hugenholtz, P., Macarie, H., Olmos, A. and Monroy, O. 2011. Multiple syntrophic
interactions in a terephthalate-degrading methanogenic consortium. The ISME
Journal, 5 (1): 122-130.
166
Madavan, N. K. 2002. Multi-objective optimization using a Pareto differential evolution
approach. Paper presented at the Proceedings of the Congress on Evolutionary
Computation 2002 (CEC‘2002), 1145–1150.
Madukasi, E. I. and Zhang, G. M. 2010. A two bioreactor treating brewery wastewater.
Advanced Materials Research, 113 - 116: 1138-1142.
Manhokwe, S., Parawira, W. and Tekere, M. 2009. An evaluation of a mesophilic reactor for
treating wastewater from a Zimbabwean potato-processing plant. African Journal of
Environmental Science and Technology, 3 (4): 091-096.
Maeder, D. L., Anderson, I., Brettin, T. S., Bruce, D. C., Gilna, P., Han, C. S., Lapidus, A.,
Metcalf, W. W., Saunders, E. and Tapia, R. 2006. The Methanosarcina barkeri
genome: comparative analysis with Methanosarcina acetivorans and Methanosarcina
mazei reveals extensive rearrangement within Methanosarcinal genomes. Journal of
Bacteriology, 188 (22): 7922-7931.
Mariani, V. C., Barbosa de Lima, A. G. and dos Santos Coelho, L. 2008. Apparent thermal
diffusivity estimation of the banana during drying using inverse method. Journal of
Food Engineering, 85 (4): 569-579.
Martinez, E., Marcos, A., Al-Kassir, A., Jaramillo, M. A. and Mohamad, A. A. 2012.
Mathematical model of a laboratory-scale plant for slaughterhouse effluents
biodigestion for biogas production. Applied Energy, 95 (0): 210-219.
Mata-Alvarez, J., Macé, S. and Llabrés, P. 2000. Anaerobic digestion of organic solid wastes.
An overview of research achievements and perspectives. Bioresource Technology, 74
(1): 3-16.
Mata, T. M., Melo, A. C., Simoes, M. and Caetano, N. S. 2012. Parametric study of a
brewery effluent treatment by microalgae Scenedesmus obliquus. Bioresource
Technology, 107: 151–158.
Maya-Altamira, L., Baun, A., Angelidaki, I. and Schmidt, J. E. 2008. Influence of wastewater
characteristics on methane potential in food processing industry wastewaters. Water
Research, 42: 2195-2203.
McHugh, S., Carton, M., Mahony, T. and O'Flaherty, V. 2003a. Methanogenic population
structure in a variety of anaerobic bioreactors. FEMS Microbiology Letters, 219: 297–
304.
McHugh, S., O‘Reilly, C., Mahony, T., Colleran, E. and O‘Flaherty, V. 2003b. Anaerobic
granular sludge bioreactor technology. Reviews in Environmental Science and
Biotechnology, 2: 225–245.
167
McKeown, R. M., Scully, C., Enright, A.-M., Chinalia, F. A., Lee, C., Mahony, T., Collins,
G. and O'Flaherty, V. 2009. Psychrophilic methanogenic community development
during long-term cultivation of anaerobic granular biofilms. The ISME Journal, 3
(11): 1231-1242.
Melamane, X. L. 2007. Treatment of wine distillery wastewaters by high rate anaerobic
digestion and submerged membrane systems. Rhodes University.
Meuer, J., Kuettner, H. C., Zhang, J. K., Hedderich, R. and Metcalf, W. W. 2002. Genetic
analysis of the archaeon Methanosarcina barkeri Fusaro reveals a central role for Ech
hydrogenase and ferredoxin in methanogenesis and carbon fixation. Proceedings of
the National Academy of Sciences, 99 (8): 5632-5637.
Mezura-Montes, E., Reyes-Sierra, M. and Coello, C. 2008. Multi-objective optimization
using differential evolution: a survey of the state-of-the-art. In: Chakraborty UK,
editor. Berlin: Springer; 2008. p.173–96[ISBN978-3-540-68827-3].
Miksch, K. and Beata, K. 2012. Distribution of Extracellular Polymeric Substances and their
Role in Aerobic Granule Formation. Available:
http://www.degruyter.com/view/j/cpe.2012.33.issue-4/v10176-012-0057-3/v10176-
012-0057-3.xml (Accessed date: 2013-12-02).
Mirzoyan, N., Parnes, S., Singer, A., Tal, Y., Sowers, K. and Gross, A. 2008. Quality of
brackish aquaculture sludge and its suitability for anaerobic digestion and methane
production in an upflow anaerobic sludge blanket (UASB) reactor. Aquaculture, 279
(1-4): 35-41.
Mirzoyan, N., Tal, Y. and Gross, A. 2010. Anaerobic digestion of sludge from intensive
recirculating aquaculture systems. Review Aquaculture, 306: 1–6.
Mohebbi, M., Barouei, J., Akbarzadeh-T, M. R., Rowhanimanesh, A. R., Habibi-Najafi, M.
B. and Yavarmanesh, M. 2008. Modeling and optimization of viscosity in enzyme-
modified cheese by fuzzy logic and genetic algorithm. Computers and Electronics in
Agriculture, 62 (2): 260-265.
Mshandete, A., Bjornsson, L., Kivaisi, A. K., Rubindamayugi, S. T. and Mattiasson, B. 2005.
Enhancement of anaerobic batch digestion of sisal pulp waste by mesophilic aerobic
pre-treatment. Water Research, 39: 1569–1575.
Mu, S. J., Zeng, Y., Wu, P., Lou, S. J. and Tartakovsky, B. 2008. Anaerobic digestion model
no. 1-based distributed parameter model of an anaerobic reactor: I. Model
development. Bioresource Technology, 99 (9): 3665-3675.
Muda, K., Aris, A., Salim, M. R., Ibrahim, Z., Loosdrecht, M. C. M. v., Ahmad, A. and
Nawahwi, M. Z. 2011. The effect of hydraulic retention time on granular sludge
biomass in treating textile wastewater. Water Research, 45: 4711-4721.
168
Mudhoo, A. and Kumar, S. 2013. Effects of heavy metals as stress factors on anaerobic
digestion processes and biogas production from biomass. International Journal
Science Environment and Technology, 10: 1383–1398.
Mukhopadhyay, D. M., Balitanas, M. O., A, A. F., Jeon, S.-H. and Bhattacharyya, D. 2009.
Genetic algorithm: A tutorial review. International Journal of Grid and Distributed
Computing, 2 (3): 25-32.
Mumme, J., Linke, B. and Tolle, R. 2010. Novel upflow anaerobic solid-state (UASS)
reactor. Bioresource Technology, 101: 592–599.
Nacheva, P. M., Pantoja, M. R. and Serrano, E. A. 2011. Treatment of slaughterhouse
wastewater in upflow anaerobic sludge blanket reactor. Water Science and
Technology, 63 (5): 877-884.
Nadais, H., Barbosa, M., Capela, I., Arroja, L., Ramos, C. G., Grilo, A., A, S., Jorge, S. and
Leitao, H. 2011. Enhancing wastewater degradation and biogas production by
intermittent operation of UASB reactors. Energy, 36 2164-2168.
Najafpour, G. D., Zinatizadeh, A. A. L., Mohamed, A. R., Hasnain, I. M. and
Nasrollahzadeh, H. 2006. High-rate anaerobic digestion of palm oil mill effluent in an
upflow anaerobic sludge-fixed film bioreactor. Process Biochemistry, 41: 370-379.
Nakasaki, K., Kwon, S. H. and Ikeda, H. 2013. Identification of microorganisms in the
granules generated during methane fermentation of the syrup wastewater produced
while canning fruit. Process Biochemistry, 48 (5): 912-919.
Nariman-Zadeh, N., Salehpour, M., Jamali, A. and Haghgoo, E. 2010. Pareto optimization of
a five-degree of freedom vehicle vibration model using a multi-objective uniform-
diversity genetic algorithm (MUGA). Engineering Applications of Artificial
Intelligence, 23 (4): 543-551.
Narra, M., Balasubramanian, V., Mehta, H., Dixit, G., Madamwar, D. and Shah, A. R. 2014.
Performance evaluation of anaerobic hybrid reactors with different packing media for
treating wastewater of mild alkali treated rice straw in ethanol fermentation process.
Bioresource Technology, 152: 59-65.
National Solid Waste Association of India (NSWAI-ENVIS) (2007). Urban Municipal Solid
Waste Management. India, New Delhi. www.nswai.com.
Nelson, M. C., Morrison, M. and Yu, Z. 2011. A meta-analysis of the microbial diversity
observed in anaerobic digesters. Bioresource Technology, 102 (4): 3730-3739.
Nercessian, O., Bienvenu, N., Moreira, D., Prieur, D. and Jeanthon, C. 2005. Diversity of
functional genes of methanogens, methanotrophs and sulfate reducers in deep‐sea
hydrothermal environments. Environmental Microbiology, 7 (1): 118-132.
169
Nery, V. D., Damianovic, M. H. Z. and Barros, F. G. 2001. The use of upflow anaerobic
sludge blanket reactors in the treatment of poultry slaughterhouse wastewater. Water
Science and Technology, 44 (4): 83-88.
Nettmann, E., Bergmann, I., Mundt, K., Linke, B. and Klocke, M. 2008. Archaea diversity
within a commercial biogas plant utilizing herbal biomass determined by 16Sr DNA
and mrcA analysis. Journal of Applied Microbiology, 105: 1835-1850.
Nigel, K. and Sneeringer, S. 2011. Carbon Emissions, Renewable Electricity and Profits:
Comparing Alternative Policies to Promote Anaerobic Digesters on Dairies. Paper
presented at the Selected paper prepared for presentation at the Annual Meeting of the
AAEA, Pittsburgh, Pennsylvania, July 24-26, 2011.
Nissilä, M. E., Li, Y.-C., Wu, S.-Y., Lin, C.-Y. and Puhakka, J. A. 2012. Hydrogenic and
methanogenic fermentation of birch and conifer pulps. Applied Energy, 100 (0): 58-
65.
Niu, Q., Qiao, W., Qiang, H. and Li, Y.-Y. 2013. Microbial community shifts and biogas
conversion computation during steady inhibited and recovered stages of thermophilic
methane fermentation on chicken manure with a wide variation of ammonia.
Bioresource Technology, 146: 223-233.
Nizami, A.-S. and Murphy, J. D. 2010. What type of digester configurations should be
employed to produce biomethane from grass silage? Renewable and Sustainable
Energy Reviews, 14: 1558–1568.
Nobu, M. K., Narihiro, T., Tamaki, H., Qiu, Y.-L., Sekiguchi, Y., Woyke, T., Goodwin, L.,
Davenport, K. W., Kamagata, Y. and Liu, W.-T. 2014. Draft genome sequence of
Syntrophorhabdus aromaticivorans strain UI, a mesophilic aromatic compound-
degrading syntrophy. Genome announcements, 2 (1): e01064-01013.
Novak, D., Franke-Whittle, I. H., Pirc, E. T., Jerman, V., Insam, H., Logar, R. M. and Stres,
B. 2013. Biotic and abiotic processes contribute to successful anaerobic degradation
of cyanide by UASB reactor biomass treating brewery waste water. Water Research,
47 (11): 3644-3653.
Ochieng, A. A., Ogadab, T., Sisenda, W. C. and P, W. 2003. Brewery wastewater treatment
in a fluidized bed bioreactor. Journal of Hazardous Material B, 90: 311–321.
Okonkwo, P. C., Aderemi, B. O. and Okoli, C. S. 2013. Factors Affecting Biogas Production
during Anaerobic Decomposition of Brewery effluent- wastewater in a Fluidized Bed
Digester. Journal of Environment and Earth Science, 3 (8): 2224-3216. ISSN 2224-
3216 (Paper) ISSN 2225-0948.
170
Oktem, Y. and Tufekei, N. 2006. Treatment of brewery wastewater by pilot scale upflow
anaerobic brewery wastewater by pilot scale upflow anaerobic sludge blanket reactor
in mesophilic temperature. Journal of Scientific and Industrial Research, 65: 248-257.
Olofintoye, O., Adeyemo, J. and Otieno, F. 2014. A Combined Pareto Differential Evolution
Approach for Multi-objective Optimization. In: Schuetze, O., Coello Coello, C. A.,
Tantar, A.-A., Tantar, E., Bouvry, P., Moral, P. D. and Legrand, P. eds. EVOLVE - A
Bridge between Probability, Set Oriented Numerics, and Evolutionary Computation
III. Springer International Publishing, 213-231. Available:
http://dx.doi.org/10.1007/978-3-319-01460-9_10.
Onodera, T., Sase, S., Choeisai, P., Yoochatchaval, W., Sumino, H., Yamaguchi, T., Ebie, Y.,
Xu, K., Tomioka, N., Mizuochi, M. and Syutsubo, K. 2013. Development of a
treatment system for molasses wastewater: The effects of cation inhibition on the
anaerobic degradation process. Bioresource Technology, 131 (0): 295-302.
Ouboter, M. R. L., EcK, B. T. M. V., Gils, J. A. G. V., Sweerts, J. P. and Villars, M. T. 1998.
Water quality modeling of the western Scheldt. Hydrobiologia 366: 129–142.
Parawira, W., Kudita, I., Nyandoroh, M. G. and Zvauya, R. 2005. A study of industrial
anaerobic treatment of opaque beer brewery wastewater in a tropical climate using a
full-scale UASB reactor seeded with activated sludge. Process Biochemistry, 40: 593–
599.
Park, E.-J., Chang, H.-W., Kim, K.-H., Nam, Y.-D., Roh, S. W. and Bae, J.-W. 2009.
Application of quantitative real-time PCR for enumeration of total bacterial, archaeal,
and yeast populations in kimchi. The Journal of Microbiology, 47 (6): 682-685.
Parsamehr, M. 2012. Modeling and analysis of a UASB reactor. Master of Science
Environmental Engineering, Master of Science Thesis. Luleå University of
Technology.
Phiri, O., Mumba, P., Moyo, B. and Kadewa, W. 2005. Assessment of the impact of
industrial effluents on water quality of receiving rivers in urban areas of Malawi.
International Journal of Environmental Science and Technology, 2 (3): 237-244.
Pierreval, H., Caux, C., Paris, J. L. and Viguier, F. 2003. Evolutionary approaches to the
design and organization of manufacturing systems. Computers and Industrial
Engineering, 44 (3): 339-364.
Pitryuk, A. V. and Pusheva, M. A. 2001. Different ionic specificities of ATP synthesis in
extremely alkaliphilic sulphate-reducing and acetogenic bacteria. Microbiology, 41
83-89.
Poh, P. E. and Chong, M. F. 2009. Development of anaerobic digestion methods for palm oil
mill effluent (POME) treatment. Bioresource Technology, 100: 1-9.
171
Pontes, R. F. F. and Pinto, J. M. 2006. Analysis of integrated kinetic and flow models for
anaerobic digesters. Chemical Engineering Journal, 122 (1–2): 65-80.
Price, K. V., Storn, R. M. and Lampinen, J. A. 2005. Differential Evolution APractical
Approach to Global Optimization. First edition ed. Springer-Verlag Berlin
Heidelberg.
Pycke, B., Etchebehere, C., Van de Caveye, P., Negroni, A., Verstraete, W. and Boon, N.
2011. A time-course analysis of four full-scale anaerobic digesters in relation to the
dynamics of change of their microbial communities. Water Science and Technology,
63 (4): 769 - 775.
Qiao, J.-T., Qiu, Y.-L., Yuan, X.-Z., Shi, X.-S., Xu, X.-H. and Guo, R.-B. 2013. Molecular
characterization of bacterial and archaeal communities in a full-scale anaerobic
reactor treating corn straw. Bioresource Technology, 143: 512-518.
Qiao, W., Peng, C., Wang, W. and Zhang, Z. 2011. Biogas production from supernatant of
hydrothermally treated municipal sludge by upflow anaerobic sludge blanket reactor.
Bioresource Technology, 102 (21): 9904-9911.
Qiu, Y.-L., Hanada, S., Ohashi, A., Harada, H., Kamagata, Y. and Sekiguchi, Y. 2008.
Syntrophorhabdus aromaticivorans gen. nov., sp. nov., the first cultured anaerobe
capable of degrading phenol to acetate in obligate syntrophic associations with a
hydrogenotrophic methanogen. Applied and Environmental Microbiology, 74 (7):
2051-2058.
Rajagopal, R., Ganesh, R., Escudie, R., Mehrotra, I., Kumar, P., Thanikal, J. V. and Torrijos,
M. 2009. High rate anaerobic filters with floating supports for the treatment of
effluents from small-scale agro-food industries. Desalination and Water Treatment, 4:
183–190.
Rajasundari, K. and Murugesan, R. 2011. Decolourization of distillery waste water–role of
microbes and their potential oxidative enzymes (Review). Journal of Applied
Environmental and Biological Sciences, 1 (4): 54-68.
Rajeshwari, K. V., Balakrishnan, M., Kansal, A., Lata, K. and Kishore, V. V. N. 2000. State-
of-the-art of anaerobic digestion technology for industrial wastewater treatment.
Renewable and Sustainable Energy Reviews, 4 (2): 135-156.
Rajput, V. S., Sharma, A. K., Ranjan, R. K. and Singh, S. 2012. Recovery of energy from
waste generated in biogas power plant. International Journal of Scientific Research
Engineering and Technology, 1 (5): 068-072.
Ralph, M. and Dong, G. J. 2010. Environmental Microbiology Second. A John Wiley &
Sons, Inc., Publication.
172
Rao, A. G., Reddy, T. S. K., Prakash, S. S., Vanajakshi, J., Joseph, J. and Sarma, P. N. 2007.
pH regulation of alkaline wastewater with carbon dioxide: A case study of treatment
of brewery wastewater in UASB reactor coupled with absorber. Bioresource
Technology, 98 (11): 2131-2136.
Raposo, M. F. J., Oliveira, S. E., Castro, P. M., Bandarra, N. M. and Morais, R. M. 2010. On
the utilization of microalgae for brewery effluent treatment and possible applications
of the produced biomass. Journal-Institute of Brewing, 116 (3): 285–292.
Raskin, L., Rittmann, B. E. and Stahl, D. A. 1996. Competition and coexistence of sulfate-
reducing and methanogenic populations in anaerobic biofilms. Applied and
Environmental Microbiology 62: 3847–3857.
Raskin, L., Stromley, J. M., Rittmann, B. E. and Stah, D. A. 1994. Group specific 16S rRNA
hybridization probes to describe natural communities of methanogens. Applied and
Environmental Microbiology, 60: 1232-1240.
Rastogi, G., Ranade, D., Yeole, T., Patole, M. and Shouche, Y. 2008. Investigation of
methanogen population structure in biogas reactor by molecular characterization of
methyl-coenzyme M reductase A (mcrA) genes. Bioresource Technology, 99 (13):
5317 - 5326.
Reungsang, A., Pattra, S. and Sittijunda, S. 2012. Optimization of key factors affecting
methane production from acidic effluent coming from the sugarcane juice hydrogen
fermentation process. Energies, 5: 4746-4757.
Rincón, B., Borja, R., González, J. M., Portillo, M. C. and Saiz-Jiménez, C. S. 2008.
Influence of organic loading rate and hydraulic retention time on the performance,
stability and microbial communities of one-stage anaerobic digestion of two-phase
olive mill solid residue. Biochemical Engineering Journal, 40: 253–261.
Rodríguez-Fernández, M., Balsa-Canto, E., Egea, J. A. and Banga, J. R. 2007. Identifiability
and robust parameter estimation in food process modeling: Application to a drying
model. Journal of Food Engineering, 83 (3): 374-383.
Romero García, L.I. 1991. Development of a general mathematical model for fermentation
processes: kinetics of anaerobic digestion. Doctoral Thesis. University of Cádiz.
ISBN, 84-7786-109-9.
Ronen, M., Shabtai, Y. and Guterman, H. 2002. Optimization of feeding profile for a fed-
batch bioreactor by an evolutionary algorithm. Journal of Biotechnology, 97 (3): 253-
263.
Rosenwinkel, K.-H., Austermann-Haun, U. and Meyer, H. 2005. Industrial Wastewater
Sources and Treatment Strategies. Environmental Biotechnology: Wiley-VCH Verlag
GmbH & Co. KGaA,
173
Ross, W. R. 1989. Anaerobic treatment of industrial effluent in South Africa. Water SA, 15
(4): 231-246.
Ross, W. R. and Louw, J. M. 1987. Monitoring and control of anaerobic digestion. Water SA,
13: 193-196.
Rüffer, H., Rosenwinkel, K.-H. E., Industrieabwasserreinigung, T. and München, W. 1991.
R. Oldenbourg Verlag.
Ryan, P., Forbes, C., McHugh, S., O'Reilly, C., Fleming, G. T. A. and Colleran, E. 2010.
Enrichment of acetogenic bacteria in high rate anaerobic reactors under mesophilic
and thermophilic conditions. Water Research, 44 (14): 4261-4269.
Saitou, N. and Nei, M. 1987. The neighbor-joining method: A new method for reconstructing
phylogenetic trees. Molecular Biology and Evolution, 4: 406-425.
Salvador, A. F., Cavaleiro, A. J., Sousa, D. Z., Alves, M. M. and Pereira, M. A. 2013.
Endurance of methanogenic Archaea in anaerobic bioreactors treating oleate-based
wastewater. Applied Microbiology and Biotechnology, 97 (5): 2211-2218.
Sambrook, J. and Russell, D. W. 2001. Gel electrophoresis of DNA and pulsed field agarose
gel electrophoresis, in: J. Sambrook, D.W. Russell (Eds.), Molecular cloning, a
laboratory manual, third ed, Cold Spring Harbour Laboratory Press, New York, NY,
USA, 5.4–5.17.
Sánchez, E., Borja, R., Travieso, L., Mart n, A. and Colmenarejo, M. F. 2005. Effect of
organic loading rate on the stability, operational parameters and performance of a
secondary upflow anaerobic sludge bed reactor treating piggery waste. Bioresource
Technology, 96 (3): 335-344.
Sánchez, J. B., Quiroga, J. M. A. and Oviedo, M. D. C. 2006. Use of microbial activity
parameters for determination of a biosolid stability index. Bioresource Technology,
97: 562–568.
Sareen, R. and Gupta, S. K. 1995. Multi objective optimization of an industrial semi batch
nylon 6 reactor. Journal of Applied Polymer Science, 58 (13): 2357–2371.
Sasaki, K., Morita, M., Hirano, S.-i., Ohmura, N. and Igarashi, Y. 2011. Decreasing ammonia
inhibition in thermophilic methanogenic bioreactors using carbon fiber textiles.
Applied Microbiology and Biotechnology, 90 (4): 1555-1561.
Sawayama, S., Tada, C., Tshukahara, K. and Yagishita, T. 2004. Effect of ammonium
addition on methanogenic community in a fluidized bed anaerobic digestion. Journal
of Bioscience and Bioengineering, 97: 65-70.
174
Schink, B. 2002. Synergistic interactions in the microbial world. Antonie Van Leeuwenhoek,
81: 257–261.
Seghezzo, L., Zeeman, G., van Lier, J. B., Hamelers, H. V. M. and Lettinga, G. 1998. A
review: The anaerobic treatment of sewage in UASB and EGSB reactors. Bioresource
Technology, 65 (3): 175-190.
Sekiguchi, Y., Kamagata, Y., Nakamura, K., Ohashi, A. and Harada, H. 1999. Fluorescence
in-situ hybridization using 16S rRNA-targeted oligonucleotides reveals localization of
methanogens and selected uncultured bacteria in mesophilic and thermophilic sludge
granules. Applied and Environmental Microbiology, 65 (3): 1280-1288.
Sekiguchi, Y., Kamagata, Y., Syutsubo, K., Ohashi, A., Harada, H. and Nakamura, K. 1998.
Phylogenetic diversity of mesophilic and thermophilic granular sludges determined by
16s rRNA gene analysis. Microbiology, 144: 2655-2665.
Sekiguchi, Y., Takahashi, H., Kamagata, Y., Ohashi, A. and Harada, H. 2001. In-situ
detection, isolation, and physiological properties of a thin filamentous microorganism
abundant in methanogenic granular sludges: a novel isolate affiliated with a clone
cluster, the green non-sulfur bacteria, subdivision I. Applied and Environmental
Microbiology, 67 (12): 5740-5749.
Sendrescu, D. 2013. Parameter identification of anaerobic wastewater treatment bioprocesses
using particle swarm optimization. Mathematical Problems in Engineering, 2013: 8.
Senturk, E., Ýnce, M. and Engin, G. O. 2013. Assesment of kinetic parameters for
thermophilic anaerobic contact reactor treating food-processing wastewater.
International Journal of Environmental Research, 7 (2): 293-302.
Shabangu, S. 2004. White Paper on the Renewable Energy Policy of the Republic of South
Africa. South Africa: Deputy Minister of Minerals and Energy.
Shaheen, H. I., Rashed, G. I. and Cheng, S. J. 2009. Application of differential evolution
algorithm for optimal location and parameters setting of UPFC considering power
system security. European Transactions on Electrical Power, 19 (7): 911-932.
Shao, X., Peng, D., Teng, Z. and Ju, X. 2008. Treatment of brewery wastewater using
anaerobic sequencing batch reactor (ASBR). Bioresource Technology, 99 (8): 3182-
3186.
Shen, P., Zhang, J., Zhang, J., Jiang, C., Tang, X., Li, J., Zhang, M. and Wu, B. 2013.
Changes in microbial community structure in two anaerobic systems to treat bagasse
spraying wastewater with and without addition of molasses alcohol wastewater.
Bioresource Technology, 131 (0): 333-340.
175
Shlimon, A. G., Friedrich, M. W., Niemann, H., Ramsing, N. B. and Finster, K. 2004.
Methanobacterium aarhusense sp. nov., a novel methanogen isolated from a marine
sediment (Aarhus Bay,Denmark). International Journal of Systematic and
Evolutionary Microbiology, 54 (3): 759-763.
Siddique, T., Penner, T., Klassen, J., Nesbø, C. and Foght, J. M. 2012. Microbial
communities involved in methane production from hydrocarbons in oil sands tailings.
Environmental Science and Technology, 46 (17): 9802-9810.
Simate, G. S., Cluett, J., Iyuke, S. E., Musapatika, E. T., Ndlovu, S., Walubita, L. F. and
Alvarez, A. E. 2011. The treatment of brewery wastewater for reuse: State of the art.
Desalination, 273 (2-3): 235-247.
Sipma, J., Osuna, M. B., Emanuelsson, M. A. E. and Castro, P. M. L. 2010. Biotreatment of
industrial wastewaters under transient-state conditions: process stability with
fluctuations of organic load, substrates, toxicants, and environmental parameters.
Critical Reviews in Environmental Science and Technology, 40, 147-197, DOI:
10.1080/10643380802039329.
Singh, K. S. and Viraraghavan, T. 2000. Performance of UASB reactor at 6 to 32 ˚C in
municipal wastewater treatment. Water Quality Research Journal of Canada, 35 (1):
113-124.
Singh, S. P. and Prerna, P. 2009. Review of recent advances in anaerobic packed-bed biogas
reactors. Renewable and Sustainable Energy Reviews, 13: 1569–1575.
Sinha, S., Bose, P., Jawed, M., John, S. and Tare, V. 2002. Application of neural network for
simulation of upflow anaerobic sludge blanket (UASB) reactor performance.
Biotechnology and Bioengineering, 77 (7): 806-814.
Smith, J. M., Castro, H. and Ogram, A. 2007. Structure and function of methanogens along a
short-term restoration chronosequence in the Florida everglades [down-pointing small
open triangle. Applied and Environmental Microbiology, 73 (13): 4135–4141.
Smith, P. H. 1966. The microbial ecology of sludge methanogenesis. Developments in
Industrial Microbiology, 7: 156-161.
Song, H., Li, Z., Du, B., Wang, G. and Ding, Y. 2012. Bacterial communities in sediments of
the shallow Lake Dongping in China. Journal of Applied Microbiology, 112 (1): 79-
89.
Soons, Z. I. T. A., M. Streefland, G. van Straten and A.J.B. van Boxtel. 2008. Assessment of
near infrared and ―software sensor‖ for biomass monitoring and control.
Chemometrics and Intelligent Laboratory Systems, 94: 166–174.
176
Sousa, D. Z., Salvador, A. F., Ramos, J., Guedes, A. P., Barbosa, S., Stams, A. J. M., Alves,
M. M. and Pereira, M. A. 2013. Activity and viability of methanogens in anaerobic
digestion of unsaturated and saturated long-chain fatty acids. Applied and
Environmental Microbiology, 79 (14): 4239-4245.
South African Breweries plc. (SAB) (2001). Corporate Citizenship Review.
Speece, R. E., Parkin, G. F. and Gallagher, D. 1983. Nickel stimulation of anaerobic
digestion. Water Research, 17 (6): 677–683.
Sponza, D. and Uluköy, A. 2008. Kinetic of carbonaceous substrate in an upflow anaerobic
sludge blanket (UASB) reactor treating 2,4 dichlorophenol (2,4 DCP). Journal of
Environmental Management, 86: 121–131.
Srinivas, N. and Deb, K. 1994. Multi objective function optimization using NSGA.
Evolutionary programming, 2.3: 221–248.
Srisertpol, J., Srinakorn, P., Kheawnak, A., Chamniprasart, K. and Srikaew, A. 2010.
Estimation dynamical model of an anaerobic digestion of shrimp culture pond
sediment in a biogas process using genetic algorithm. Paper presented at the
Proceedings of the 9th WSEAS international conference on System science and
simulation in engineering. Japan.1938982: World Scientific and Engineering Academy
and Society (WSEAS), 449-453.
Stafford, W., Cohen, B., Pather-Elias, S., von Blottnitz, H., van Hille, R., Harrison, S. T. L.
and Burton, S. G. 2013. Technologies for recovery of energy from wastewaters:
Applicability and potential in South Africa. Journal of Energy in Southern Africa, 24:
00-00.
Stahl, D. A. and Amann, R. 1991. Development and application of nucleic acid probes, p.
207–248. In E. Stackebrandt and M. Goodfellow (ed.). Nucleic acid techniques in
bacterial systematics, vol. 8. John Wiley & Sons, London, England.
Steinberg, L. M. and Regan, J. M. 2009. mcrA-targeted real-time quantitative PCR method to
examine methanogen communities. Applied and Environmental Microbiology, 75
(13): 4435–4442.
Steinhaus, B., Garcia, M. L., Shen, A. Q. and Angenent, L. T. 2007. A portable anaerobic
microtank reveals optimum growth conditions for the methanogen Methanosaeta
concilii. Applied and Environmental Microbiology, 73: 1653–1658.
Sternenfels, U. M. D. C. 2012. Compost Physicochemical Characteristics Influencing
Methane Biofiltration. Degree of Doctor of Philosophy, University of Calgary.
177
Storn, R. and Price, K. 1995. Differential evolution: a simple and efficient adaptive scheme
for global optimization over continuous spaces, . Technical Report TR-95-012,
International Computer Science Institute, Berkeley, USA.
Sundberg, C., Al-Soud, W. A., Larsson, M., Alm, E., Yekta, S. S., Svensson, B. H., Sørensen,
S. J. and Karlsson, A. 2013. 454 pyrosequencing analyses of bacterial and archaeal
richness in 21 full-scale biogas digesters. FEMS Microbiology Ecology, 85 (3): 612-
626.
Supaphol, S., Jenkins, S., Intomo, P., Waite, I. and O'Donnell, A. 2011. Microbial community
dynamics in mesophilic anaerobic co-digestion of mixed waste. Bioresource
Technology, 102 (5): 4021 - 4027.
Sykes, R. M. 1995. Biological water treatment processes, in The Civil Engineering
Handbook, Chen, W.F. (Ed). (CRC Press LLC, New York).
Tabatabaei, M., Sulaiman, A., Nikbakht, A. M., Yusof, N. and Najafpour, G. 2011.
Influential parameters on biomethane generation in anaerobic wastewater treatment
plants.
Tabatabaei, M., Zakaria, M. R., Rahim, R. A., Wright, A.-D. G., Yoshihito, S., Abdullah, N.,
Kenji, S., Ikeno, S., Masatsugu, M., Nakamura, K., Sulaiman, A. and Hassan, M. A.
2009. PCR-based DGGE and FISH analysis of methanogens in an anaerobic closed
digester tank for treating palm oil mill effluent. Electronic Journal of Biotechnology
[online]. July 15, 2009.vol. 12, no. 3. Available from
Internet:http://www.ejbiotechnology.cl/content/vol12/issue3/full/4/index.html:10.222
5/vol12-issue3-full text-4.
Tai, C. S., Singh, K. S. and Grant, S. R. 2006. Combined removal of carbon and nitrogen in
an integrated UASB-jet loop reactor bioreactor system. Journal of Environmental
Engineering, 132: 624-637.
Talbot, G., Topp, E., Palina, M.F. and Masse, D.I. 2008. Evaluation of molecular methods
used for establishing the interactions and functions of microorganisms in anaerobic
bioreactors-Review. Water Research 42: 513-537.
Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M. and Kumar, S. 2011. MEGA5:
Molecular evolutionary genetics analysis using maximum likelihood, evolutionary
distance, and maximum Parsimony methods. Molecular Biology and Evolution, 28:
2731-2739.
Tan, C., Ma, F. and Qiu, S. 2013. Impact of carbon to nitrogen ratio on nitrogen removal at a
low oxygen concentration in a sequencing batch biofilm reactor. Water science and
technology: Journal of the International Association on Water Pollution Research, 67
(3): 612-618.
178
Tarafder, A., Rangaiah, G. P. and Ray, A. K. 2005. Multi-objective optimization of an
industrial styrene monomer manufacturing process. Chemical Engineering Science,
60 (2): 347–363.
Tauseef, S. M., Abbasi, T. and Abbasi, S. A. 2013. Energy recovery from wastewaters with
high-rate anaerobic digesters. Renewable and Sustainable Energy Reviews, 19 (0):
704-741.
Tay, J. H. and Zhang, X. 1999. "Neural fuzzy modeling of anaerobic biological wastewater
treatment systems‖ ASCE Journal of Environmental Engineering, 125 (12): 149-
1159.
The Centre for Sustainable Environmental Sanitation (CSES) (2009). Opportunities for
German know-how and CDM application'-The Chinese Biomass Sector. CDM
Perspective in China. T. Z. G. G. Deutsche Gesellschaft für, Beijing Office,
Sunflower Tower 1100,37 Maizidian Street, Chaoyang District, 100125 Beijing, PR
China and www.gtz.de. Beijing, The University of Science and Technology, Beijing.
http://www.jiko-bmu.de/files/inc/application/pdf/gtz-china_cdm_sector_study-
waste_water_0907.pdf. (Accessed date: 2012-08-10)
Thomas, H. C. 2010. Solid waste technology and management Anaerobic digestion: process.
A John Wiley and Sons, Limited Publication, Volume 2.
Thorin, E., Nordlander, E., Lindmark, J., Dahlquist, E., Yan, J. and Fdhila, R. B. 2012.
Modeling of the biogas production process- a review. Paper presented at the
International Conference on Applied Energy (ICAE), Paper ID: ICAE2012- A10732,
Jul 5-8, 2012, Suzhou, China
Tiwari, M., Guha, S., Harendranath, C. S. and Tripathi, S. 2006. Influence of extrinsic factors
on granulation in UASB reactor. Applied Microbiology and Biotechnology, 71: 145–
154.
Torkian, A., Eqbali, A. and Hashemian, S. J. 2003. The effect of organic loading rate on the
performance of UASB reactor treating slaughterhouse effluent. Resources,
Conservation and Recycling, 40 (1): 1-11.
Traversi, D., Capone, C., Villa, S., Valeria, R., Pietrangeli, B. and Gilli, G. 2014. Assessing
archeal indicators of performance by RT-qPCR methods during anaerobic co-
digestion of organic wastes. BioEnergy Research: 1-8.
Traversi, D., Villa, S., Acri, M., Pietrangeli, B., Degan, R. and Gilli, G. 2011. The role of
different methanogen groups evaluated by real-time QPCR as high-efficiency
bioindicators of wet anaerobic co-digestion of organic waste. AMB Express, 1 (1): 28.
179
Traversi, D., Villa, S., Lorenzi, E., Degan, R. and Gilli, G. 2012. Application of a real-time
QPCR method to measure the methanogen concentration during anaerobic digestion
as an indicator of biogas production capacity. Journal of Environmental Management,
111: 173-177.
Tsai, K.-Y. and Wang, F.-S. 2005. Evolutionary optimization with data collocation for
reverse engineering of biological networks. Bioinformatics, 21 (7): 1180–1188.
Turkdogan-Aydinol, F. I. and Yetilmezsoy, K. 2010. A fuzzy-logic-based model to predict
biogas and methane production rates in a pilot-scale mesophilic UASB reactor
treating molasses wastewater. Journal of Hazardous Materials, 182 (1-3): 460-471.
Turovskiy, I. S. and Mathai, P. K. 2006. Wastewater Sludge Processing. Wiley.
Trzcinski, A. P., Ray, M. J. and Stuckey, D. C. 2010. Performance of a three-stage membrane
bioprocess treating the organic fraction of municipal solid waste and evolution of its
archaeal and bacterial ecology. Bioresource Technology, 101 (6): 1652-1661.
Uneo, Y. and Tatara, M. 2008. Microbial population in a thermophilic packed-bed reactor for
methanogenesis from volatile fatty acids. Enzyme and Microbial Technology, 43:
302– 308
United Nations Economic Commission for Europe (UNEP). 2004. Eco-efficiency for the
Dairy Processing Industry Manual, August 2004.
Uyanik, S., Sallis, E. J. and Anderson, G. K. 2002. The effect of polymer addition on
granulation in an anaerobic baffled reactor (ABR): Part I. Process performance. Water
Research, 36: 933-942.
Van de Wijngaard, W., Creemers, J., Vogels, G. and Van der Drift, C. 1991. Methanogenic
pathways in Methanosphaera stadtmanae. FEMS Microbiology Letters, 80 (2): 207-
211.
Vasant, P. and Barsoum, N. 2009. Hybrid genetic algorithms and line search method for
industrial production planning with non-linear fitness function. Engineering
Applications of Artificial Intelligence, 22 (4-5): 767-777.
Vavilin, V., Qu, X., Mazeas, L., Lemunier, M., Duquennoi, C., He, P. and Bouchez, T. 2008.
Methanosarcina as the dominant aceticlastic methanogens during mesophilic
anaerobic digestion of putrescible waste. Antonie Van Leeuwenhoek International
Journal of General and Molecular Microbiology, 94 (4): 593 - 605.
Visser, A., Gao, Y. and Letingga, G. 1993. Effects of pH on methanogenesis and sulphate
reduction in thermophilic (55oC) UASB reactors. Bioresource Technology, 44: 113–
121.
180
Vlyssides, A., Barampouti, E. M. and Mai, S. 2008. Granulation mechanism of a UASB
reactor supplemented with iron. Anaerobe, 14 (5): 275-279.
Wang, H., Tolvanen, K., Lehtomaki, A., Puhakka, J. and Rintala, J. 2010. Microbial
community structure in anaerobic co-digestion of grass silage and cow manure in a
laboratory continuously stirred tank reactor. Biodegradation, 21: 135-146.
Wang, Y., Zhang, Y., Wang, J. and Meng, L. 2009. Effects of volatile fatty acid
concentrations on methane yield and methanogenic bacteria. Biomass and Bioenergy,
35 (5): 848-853.
Ward, J. A., Hobbs Phil, J., Holliman, P. J. and Jones, D. L. 2008. Optimisation of the
anaerobic digestion of agricultural resources. Review Bioresource Technology, 99:
7928–7940.
Watanabe, T., Asakawa, S., Nakamura, A., Nagaoka, K. and Kimura, M. 2004. DGGE
methods for analyzing 16S rDNA of methanogenic archaeal community in paddy
field soil. FEMS Microbiology Letters, 232: 143-163.
Wei, X. and Kusiak, A. 2012. Optimization of biogas production process in a wastewater
treatment plant. In: Proceedings of the 2012 Industrial and Systems Engineering
Research Conference.
Welte, C. and Deppenmeier, U. 2013. Bioenergetics and anaerobic respiratory chains of
aceticlastic methanogens. Biochimica et Biophysica Acta.
Wen, Q., Wu, Y., Cao, D., Zhao, L. and Sun, Q. 2009. Electricity generation and modeling of
microbial fuel cell from continuous beer brewery wastewater. Bioresource
Technology, 100 (18): 4171-4175.
Werner, J. J., Knights, D., Garcia, M. L., Scalfone, N. B., Smith, S., Yarasheski, K.,
Cummings, T. A., Beers, A. R., Knight, R. and Angenent, L. T. 2011. Bacterial
community structures are unique and resilient in full-scale bioenergy systems.
Proceedings of the National Academy of Sciences, 108 (10): 4158-4163.
Wiegant, W. M. 2001. Experiences and potentials of anaerobic wastewater treatment in
tropical regions Anearobic digstion of sustainable development. Farewell seminar of
Prof dr. Ir. Gatze Lettinga Wageningen - The Netherlands, EP & RC: 111-118.
Wikström, T., Nordmark, D., Pelkonen, M. and Lagerkvist, A. 2012. Fluorescent in situ
hybridization technique in anaerobic process studies. In: Lagerkvist, A. ed.
Proceedings of Abstract proceedings of 7th Intercontinental Landfill Research
Symposium Sunderbyn, Luleå, Sweden, Luleå tekniska universitet.
181
Wirth, R., Kovacs, E., Maroti, G., Bagi, Z., Rakhely, G. and Kovacs, K. 2012.
Characterization of a biogas-producing microbial community by short-read next
generation DNA sequencing. Biotechnology for Biofuels, 5 (1): 41.
Woldesenbet, Y. G., Yen, G. G. and Tessema, B. G. 2009. Constraint handling in multi-
objective evolutionary optimization. Evolutionary Computation, IEEE Transactions
on, 13 (3): 514-525.
Won, S. G. and Lau, A. K. 2011. Effects of key operational parameters on biohydrogen
production via anaerobic fermentation in a sequencing batch reactor. Bioresource
Technology, 102: 6876–6883.
Worldwide Brewing Alliance 2011. Report on environment and utilities sustainability.
Wu, J.-H., Liu, W.-T., Tseng, I.-C. and Cheng, S.-S. 2001. Characterization of microbial
consortia in a terephthalate-degrading anaerobic granular sludge system.
Microbiology, 147 (2): 373-382.
Wu, P., Zeng, Y., Mu, S., Lou, S., Tartakovsky, B. and Guiot, S. 2005. Hydraulic modelling
and axial dispersion analysis of UASB reactor. Biochemical Engineering Journal, 25:
113–123.
Yadvikaa, S., Sreekrishnanb, T. R., Kohlic, S. and Ranaa, V. 2004. Enhancement of biogas
production from solid substrates using different techniques––a review. Bioresource
Technology, 95 (1): 1–10.
Ye, W., Liu, X., Lin, S., Tan, J., Pan, J., Li, D. and Yang, H. 2009. The vertical distribution
of bacterial and archaeal communities in the water and sediment of Lake Taihu.
FEMS Microbiology Ecology, 70 (2): 263-276.
Yee, A. K. Y., Ray, A. K. and Rangiah, G. P. 2003. Multi-objective optimization of industrial
styrene reactor. Computers and Chemical Engineering, 27: 111–130.
Yetilmezsoy, K. 2012. Integration of kinetic modeling and desirability function approach for
multi-objective optimization of UASB reactor treating poultry manure wastewater.
Bioresource Technology, 118 (89-101): 118 189-101.
Yetilmezsoy, K. and Sakar, S. 2008. Development of empirical models for performance
evaluation of UASB reactors treating poultry manure wastewater under different
operational conditions. Journal of Hazardous Materials, 153 (1–2): 532-543.
Yetilmezsoy, K. and Sapci-Zengin, Z. 2009. Stochastic modeling applications for the
prediction of COD removal efficiency of UASB reactors treating diluted real cotton
textile wastewater. Stochastic Environmental Research and Risk Assessment, 23: 13-
26.
182
Yu, H. Q., Fang, H. H. P. and Tay, J. H. 2001. Enhanced sludge granulation in upflow
anaerobic sludge blanket (UASB) reactors by aluminum chloride. Chemosphere, 44
(1): 31-36.
Yu, Y., Kim, J. and Hwang, S. 2006. Use of real-time PCR for group-specific quantification
of aceticlastic methanogens in anaerobic processes: population dynamics and
community structures. Biotechnology and Bioengineering, 93 (3): 424-433.
Yu, Y., Lee, C., Kim, J. and Hwang, S. 2005. Group-specific primer and probe sets to detect
methanogenic communities using quantitative realtime polymerase chain reaction.
Biotechnology and Bioengineering, 89: 670–679.
Yuan, X., Wen, B., Ma, X., Zhu, W., Wang, X., Chen, S. and Cui, Z. 2014. Enhancing the
anaerobic digestion of lignocellulose of municipal solid waste using a microbial
pretreatment method. Bioresource Technology, 154: 1-9.
Yenigun, O., Kizilgun, F. and Yilmazer, G. 1996. Inhibition effects of zinc and copper on
volatile fatty acid production during anaerobic digestion. Environmental Technology,
17.
Zainol, N. 2012. Kinetic of Biogas Production from Banana Stem Waste.
Zandvoort, M. H., van Hullebusch, E. D., Gieteling, J. and Lens, P. N. L. 2006. Granular
sludge in full-scale anaerobic bioreactors: Trace element content and deficiencies.
Enzyme and Microbial Technology, 39 (2): 337-346.
Zeikus, J. G. 1977. The biology of methanogenic bacteria. Bacteriological Reviews, 41: 514-
541.
Zeikus, J. G. 1980. Microbial populations in bioreactors. In: D.A. Stafford, B. E. W. D. H. H.
Ed. In: Anaerobic Digestion London: Applied Science Publishers, 61-81.
Zhang, J., Zhang, Y., Quan, X. and Chen, S. 2013. Effects of ferric iron on the anaerobic
treatment and microbial biodiversity in a coupled microbial electrolysis cell (MEC) –
Anaerobic reactor. Water Research, 47 (15): 5719-5728.
Zhang, L., Sun, Y., Guo, D., Wu, Z. and Jiang, D. 2012. Molecular diversity of bacterial
community of dye wastewater in an anaerobic sequencing batch reactor. African
Journal of Microbiology Research, 6 (35): 6444-6453.
Zhang, T. and Fang, H. H. P. 2006. Applications of real-time polymerase chain reaction for
quantification of microorganisms in environmental samples. Applied Microbiology
and Biotechnology, 70: 281-289.
183
Zhao, B., Mu, Y., Dong, F., B.Ni, Zhao, J., Sheng, G., Yu, H., Li, Y. and Harada, H. 2010.
Dynamic modelling of the anaerobic reactor startup process. Industrial and
Engineering Chemistry Research, 49: 7193–7200.
Zheng, D. and Raskin, L. 2000. Quantification of Methanosaeta species in anaerobic
bioreactors using genus- and species-specific hybridization probes. Microbial
Ecology, 39: 246-262.
Zhou, H., Löffler, D. and Kranert, M. 2011. Model-based predictions of anaerobic digestion
of agricultural substrates for biogas production. Bioresource Technology, 102: 10819–
10828.
Zhou, P., Su, C., Li, B. and Qian, Y. 2006. Treatment of high-strength pharmaceutical
wastewater and removal of antibiotics in anaerobic and aerobic biological treatment
processes. Journal of Environmental Engineering, 132 (1): 129-136.
Zhou, W., Imai, T., Ukita, M., Li, F. and Yuasa, A. 2007. Effect of loading rate on the
granulation process and granular activity in a bench scale UASB reactor. Bioresource
Technology, 98 (7): 1386-1392.
Zhu, C., Zhang, J., Tang, Y., Zhengkai, X. and Song, R. 2011. Diversity of methanogenic
Archaea in a biogas reactor fed with swine feces as the mono-substrate by mcrA
analysis. Microbiological Research, 166 (1): 27–35.
Ziemiński, K. and Frąc, M. 2012. Methane fermentation process as anaerobic digestion of
biomass: Transformations, stages and microorganisms. African Journal of
Biotechnology, 11 (8): 4127-4139.
Ziganshin, A., Schmidt, T., Scholwin, F., Il‘inskaya, O., Harms, H. and Kleinsteuber, S.
2011. Bacteria and Archaea involved in anaerobic digestion of distillers grains with
solubles. Applied Microbiology and Biotechnology, 89 (6): 2039-2052.
Zvauya, R., Parawira, W. and Mawadza, C. 1994. Aspects of aerobic thermophilic treatment
of Zimbabwean traditional opaque-beer brewery wastewater. Bioresource
Technology, 48: 273–274.
184
APPENDICES
APPENDIX ONE: Analysis of variance-test (Chapter 3)
Table A1: One way ANOVA for percentage COD removal and biogas
yield during anaerobic degradation
Table Analyzed Data 1
One-way analysis of variance
P value < 0.0001
Are means signif. different? (P < 0.05) Yes
Number of groups 3
F 634.0
R squared 0.9702
Bartlett's test for equal variances
Bartlett's statistic (corrected) 8.462
P value 0.0145
Do the variances differ signif. (P < 0.05) Yes
ANOVA Table SS df MS
Treatment (between columns) 34290 2 17150
Residual (within columns) 1055 39 27.04
Total 35350 41
Post test for linear trend
Slope 34.65
R squared 0.9510
P value < 0.0001
Is linear trend significant (P < 0.05)? Yes
185
APPENDIX TWO: Publications
186
187
188